<!-- Roch Dog research paper, full text. Canonical page: https://rochstandard.com/research/dogs-in-restaurants/
     PDF: https://rochstandard.com/research/dogs-in-restaurants/dogs-in-restaurants.pdf
     DOI: https://doi.org/10.5281/zenodo.24395250
     Cite as: Bule, G. (2026). No Dogs Allowed: Why Restaurants Ban Dogs and What the Evidence Says. Roch Dog. https://rochstandard.com/research/dogs-in-restaurants/ (DOI 10.5281/zenodo.24395250)
     Publisher: Roch Dog, https://rochstandard.com. Research index: https://rochstandard.com/research/ -->
# No Dogs Allowed: Why Restaurants Ban Dogs and What the Evidence Says

**Author:** [Guise Bule](https://www.linkedin.com/in/guisebule/), CEO of [Roch Dog](https://rochdog.com)
**Subtitle:** Where the Ban Came From and What Restaurants, Bars and Hotels Should Do
**Document Reference:** RDWP-DDR-01
**Version:** 1.0
**Date:** 20th November 2026

---

## Executive Summary

In much of the world a dog cannot sit under its owner's table while the owner eats, and almost everyone assumes there is a health reason for it. We went looking for that reason. We read the food codes back to 1940, the outbreak records, and every study we could find on dogs, hygiene and food. This paper sets out what we found, for the people who run restaurants, bars, pubs and hotels.

The ban is younger and thinner than it looks. In the United States the federal model code kept animals out of the kitchen and storeroom only, until a committee rewrote it in 1962 to cover every area where food is served. The reason it gave was a single sentence, filed under housekeeping beside dust and soiled linen, and it cited no study and no outbreak. Britain never banned dogs from dining rooms at all, and much of Europe leaves the decision to the owner of the business.

The science does not fill the gap. Healthy dogs do carry germs, and dogs fed raw meat carry many more, which is a fair reason to keep any dog away from food preparation and off the crockery. But the two national reports on restaurant outbreaks in the United States do not mention a dog once, the one food regulator that has formally assessed dogs in dining areas judged the risk "very low to negligible", and two studies have found dog paws cleaner than shoes. The proven source of restaurant food poisoning is people: sick food workers feature in about four in ten outbreaks where a cause is found.

There are real reasons to manage a dog in a dining room. Some guests are allergic, some dogs misbehave, and a dog in an aisle is a hazard to a server with hot plates. Those are reasons for rules about where the dog sits and how it behaves, and none of them is a reason for a law that treats every dog as a threat to the food.

A rule that shuts people out should be able to show its evidence, and in sixty years this one has not. Nobody has measured a dining room that admits dogs against one that does not, and that study is for the health authorities who enforce the rule to commission. Meanwhile the countries now opening their dining rooms, among them Australia, the United States, Singapore, South Korea and Hong Kong, are doing so with a short list of sensible conditions, and that list is the practical heart of this paper.

---

## 1. The Rule Everyone Assumes

Ask a restaurant owner in Toronto or Chicago why a dog cannot come inside and the answer is the health code. Ask one in Paris, Munich or a Yorkshire pub and they will wonder why you are asking. The rule most people assume to be universal and scientific is neither, and the place to start is with what it actually covers.

### 1.1 Nobody is asking for the kitchen

Every food law we read keeps animals out of the rooms where food is prepared and stored, and nobody argues with that. A kitchen holds open ingredients, chopping boards, clean plates and busy people, and no dog owner expects to walk a dog through it. This paper is not about the kitchen. It is about the other room, the one with the tables, where the food arrives cooked and plated and the only thing a dog would do is lie on the floor.

That is the distinction the public debate almost always misses. The kitchen rule and the dining room rule are different rules, with different histories and very different evidence behind them, and a great deal of confusion comes from treating them as one.

*The question is not whether a dog belongs in a kitchen. It is whether a dog lying under a table is a risk to the meal on top of it.*

### 1.2 Three kinds of country

Countries fall into three groups. In the first, the law stops at the kitchen door and the dining room is the owner's decision. This is the position across the European Union, whose hygiene regulation requires businesses to keep domestic animals out of "places where food is prepared, handled or stored" and says nothing about where customers eat. Britain kept the same rule after leaving the Union, which is why a dog in a pub has never been against the law.

In the second group the law reaches the dining room. The United States model code says live animals "may not be allowed on the premises of a food establishment", and Canada and Australia took the same approach indoors. Until this year so did South Korea, and so did Hong Kong, where bringing a dog onto food premises was a criminal offence.

The third group is the second group changing its mind. Florida let its towns permit dogs on restaurant patios in 2006, Australia opened outdoor dining areas nationally in 2012, and the American model code followed in 2022. Singapore relaxed its outdoor rules in 2025, and in 2026 both South Korea and Hong Kong began letting dogs inside.

*The dining room ban is a habit of some countries, and those countries have been unpicking it for twenty years.*

### 1.3 The law shows up in our own data

We grade more than 3,000 hotels in over 70 countries, and one of the things we record is whether a guest may sit with a dog in the hotel's bar, restaurant or lounge. The pattern follows the law closely. In Britain about three hotels in four in our directory welcome a dog in the bar, and in continental Europe more than half do. In the United States it is about one in four, and in Canada it is one in twenty five.

The gap is not explained by how hotels feel about dogs. All of these hotels take dogs as guests, and many of the North American ones do it generously, with beds, bowls and treats in the room. They stop at the door of the bar because the health code tells them to. In Canada more than four hotels in ten have no bar, restaurant, lounge or terrace where a guest and a dog can sit down together, against one in ten in Britain.

*Where the law leaves the choice to the hotel, most hotels choose to let the dog in. Where it does not, the guest eats alone or the dog stays upstairs.*

---

## 2. Where the Ban Came From

We traced the American rule back through the model codes that the federal government has published for local health departments since the 1930s, and read each edition in the original. The history is short, and it turns on one change made in 1962.

### 2.1 For a quarter of a century the rule stopped at the kitchen

The United States Public Health Service began work on a model restaurant ordinance in 1934. The editions of 1940 and 1943 carry the same sentence on animals, placed in the item on storing and serving food beside the rules on flies and rodents: "No animals or fowls shall be kept or allowed in any room in which food or drink is prepared or stored."

That is a kitchen and storeroom rule. It says nothing about the room where customers sit, and it was no minor document. By the Service's own count the 1943 edition, or a local law equivalent to it, was adopted by 37 states and more than 1,100 county and city health departments. For the first quarter of a century of federal restaurant hygiene, a dog in the dining room was not the code's concern.

*The original American rule kept animals out of the kitchen and stopped there.*

### 2.2 One sentence in 1962

In 1962 the Service replaced the old ordinance with a new Food Service Sanitation Manual, drafted with an advisory committee of state sanitary engineers, academics and industry representatives. The animal rule was moved and widened. It now read: "No live birds or animals shall be allowed in any area used for the storage, preparation, or serving of food", along with the areas for washing utensils, the toilets, the staff changing rooms and the delivery vehicles, "or in any other area or facility used in the conduct of food-service establishment operations".

The word that matters is serving. The committee knew it had reached the dining room, because it wrote an exception for it in the same sentence: "guide dogs accompanying blind persons may be permitted in dining areas".

The manual gives a reason for each rule. The reason given here is one sentence long, and it appears under housekeeping, after a paragraph on dust raised by sweeping and another on soiled linen: "Live birds and animals may spread disease organisms, dust, and other contaminants." There is no study behind it, no outbreak and no figure. The manual's preface says the committee identified areas that needed research and carried some of it out. Nothing in the text suggests that animals in dining rooms were among them.

*The ban on dogs in American dining rooms rests on twelve words written under the heading of housekeeping in 1962.*

### 2.3 The reason has not changed since

Every later American code descends from that wording. Today's Food and Drug Administration Food Code states the rule more broadly still, as a ban on live animals anywhere on the premises, and its explanation is still a general statement: "Animals carry disease-causing organisms and can transmit pathogens to humans through direct and/or indirect contamination of food and food-contact surfaces." It adds that animals shed hair and may leave waste. Sixty years on, the code still cites no study of a dining room.

*In sixty years the reason has been restated several times and has never once been supported with evidence.*

### 2.4 Britain never banned the dog, and the pests explain why others did

British law took a different road. The Food Hygiene (General) Regulations 1970, in force until 1995, said only that a person handling food must not carry it together with a live animal or let one touch it. The regulations that replaced them in 1995 did not mention animals at all, only pests. The first British rule keeping a dog out of a kitchen arrived on 1 January 2006 with the European regulation, and its chapter on food rooms expressly excludes dining areas.

The clue to all of this is where the animal rules sit. In the international food hygiene code published by the Food and Agriculture Organization and the World Health Organization, animals appear in the section on pest control, in a text written for factories: "Birds and animals, other than those intended for slaughter, should be excluded from establishments." The European rule on domestic animals is the second sentence of a paragraph whose first sentence is about pests. Mexico's food hygiene standard does the same, with its rule on pets placed inside the section headed pest control.

*Dogs were never assessed as a risk to diners. They were filed with the vermin, and in some countries the filing was extended from the kitchen to the whole building.*

---

## 3. What a Healthy Dog Carries

A rule can be badly made and still be right. So the fair question is what a dog brings into a room, and the research on that is the strongest part of the case for caution. We set it out in full, because an operator deserves the whole picture and not the comfortable half.

### 3.1 Low levels, and no way to tell by looking

The largest modern survey is American. Veterinary laboratories in 36 states tested 2,422 dogs between 2012 and 2014 and found Salmonella in 2.5% of them. Nearly half of the dogs that carried it, 27 of 60, had no diarrhoea and no sign of being unwell. The authors noted that the rate has fallen over the decades, and that the dogs most likely to carry it were the ones fed raw food.

A Canadian study went further. It examined 102 dogs that visited hospital patients in Ontario, every one of them judged healthy, and tested each for eighteen organisms that can pass to people. Four dogs in five carried at least one. The commonest was Clostridium difficile, found in 58% of the dogs, and Salmonella was found in 3%.

Numbers like these need reading with care. Many of the organisms are ones people carry too, and some of those counted are rarely a cause of human illness. A Canadian government review of Campylobacter in pets found that the type dogs most often carry does not necessarily make people ill, so that headline figures overstate the danger.

*A healthy, well kept dog can carry organisms that make people ill, usually at low levels, and nothing about the dog will tell you.*

### 3.2 Raw feeding changes the picture

One finding cuts against the dog in every study that looks for it. Dogs fed raw meat carry far more. Swedish researchers compared 25 dogs fed raw meat with 25 fed dry food in the same town. They found a drug resistant form of E. coli in 13 of the raw fed dogs and one of the others, and Campylobacter in 12 against four.

It does reach people. Public health investigators in Quebec traced a cluster of drug resistant Salmonella between 2021 and 2023 to its sources. Of the twenty human cases, half were infants, and the investigators named dogs fed raw meat as the main source they could identify.

A hotel or restaurant cannot know what a guest's dog had for breakfast. That is not an argument for a ban, but it is the reason the sensible controls in section 8 apply to every dog and do not depend on trust.

*Raw fed dogs are the real hazard in this literature, and since nobody can spot one, the rules have to work for every dog.*

### 3.3 The bowl is the weak point

The clearest laboratory finding concerns dishes, and it matters to anyone thinking of offering a dog a bowl. Researchers at the University of Guelph contaminated dog bowls with Salmonella in raw meat and then tried to clean them. The bacteria were still alive in every bowl after a week. They survived a dishwasher cycle at 85 degrees in two bowls out of three, and a scrub with soap followed by a bleach soak in more than four in ten.

That experiment used raw meat, the worst case, but it sets a sound rule. A dog's bowl should not be a plate from the dining room, and it should not go through the wash with them.

*Whatever a dog eats or drinks from stays separate from what people eat from, at the table and in the wash.*

### 3.4 Carrying is not the same as passing on

For a dog under a table to make a diner ill, something has to carry what the dog has to the food. Australia's food regulator set this out when it assessed the question. Either the dog touches the food, which does not happen when the dog is on the floor, or a go between is needed, most likely a member of staff who touches the dog and then the plate. The regulator's point was that two unlikely things must both happen, and that ordinary food hygiene already guards against the second.

The American record on animals fits that reasoning. The Centers for Disease Control and Prevention counted 557 outbreaks of gut illness from contact with animals between 2009 and 2021. Dogs accounted for 31 of them, and that figure includes outbreaks spread by contaminated pet food and treats. The settings were overwhelmingly private homes and farms. Restaurants do not appear in the table at all.

*People do catch things from dogs, mostly at home and mostly from their own. The dining room is not where the record says it happens.*

---

## 4. Has a Dog in a Dining Room Made Anyone Ill?

This is the question the ban depends on, and we looked for an answer in four places: the outbreak records, the regulators, the studies that compare dogs with people, and the hospitals that admit dogs to their wards.

### 4.1 The outbreak records

The United States investigates restaurant outbreaks more thoroughly than any other country, and publishes what it finds. The most detailed report covers 800 outbreaks at 875 food establishments between 2017 and 2019, four in five of them restaurants, with the causes recorded by environmental health officers who went to the premises. A second report covers 2,677 outbreaks between 2014 and 2022. We searched both for any mention of a dog, a pet or a customer's animal as a cause. Neither contains one.

Two cautions apply. The United States bans dogs indoors, so its dining rooms have had little chance to produce such an outbreak. And outbreak reports catch clusters of illness, not the single case that nobody connects to a meal. But our search of the medical literature found no such outbreak in Europe either, where dogs have sat in restaurants, cafes and pubs for generations.

*We found no recorded outbreak, in any country, traced to a customer's dog in a dining room.*

### 4.2 The one regulator that assessed it

Only one food regulator appears to have carried out a formal risk assessment, and it did so because a health department asked an awkward question. In Australia, the Western Australian Department of Health raised a concern that the restrictions on dogs "did not appear to be based on risk". Food Standards Australia New Zealand assessed the evidence and concluded in 2012 that the risk of a diner catching a disease from a dog in an outdoor dining area was "very low to negligible". It received ninety submissions, most of them in favour, and the national ban on dogs in outdoor dining areas was lifted that year.

The assessment covered outdoor areas because that was the question asked. Its reasoning, that the dog does not touch the food and that normal hygiene controls the staff who might, does not depend on whether there is a roof overhead.

*The only regulator to test the rule against the evidence found that it could not be supported, and changed it.*

### 4.3 Paws and shoes

Two studies have compared what a dog brings into a building with what its owner brings. Researchers at Utrecht University tested the front paws of 50 dogs and the shoe soles of their 50 owners after a walk, looking for the family of bacteria used as a marker of faecal contamination. They found none on 72% of the dogs and 42% of the people, and the counts were lower on the paws. They did the study because four in five assistance dog users told them they had been refused entry somewhere, usually on grounds of hygiene.

A Slovenian team looking for Clostridium difficile in twenty households found it on 43% of shoes, 28% of slippers and 24% of dog paws. And a Swiss hospital, checking whether dogs could share a scanner with patients, found higher bacterial counts in the beards of 18 men than in the fur of 30 dogs. These are small studies and the last is little more than a curiosity. But they all point the same way, and nobody has proposed a health code for shoes.

*Every person who walks into a dining room brings the street in with them, and on the evidence so far they bring more of it than a dog does.*

### 4.4 The hospital test

Hospitals are the hardest test of all, because their patients are the people most easily harmed by infection. Many now admit visiting dogs to their wards, and the body that advises American hospitals on infection control publishes guidance on how to do it.

The most instructive study followed 96 dogs that visited hospitals and 98 that did other visiting work, testing each every two months for a year. The hospital dogs were 4.7 times as likely to pick up MRSA and 2.4 times as likely to pick up Clostridium difficile, and the ones that licked patients or took treats from them were most at risk. The infection was travelling from the people to the dogs.

*Where dogs and vulnerable people have been studied together most closely, it was the dogs that came away carrying what the people had.*

---

## 5. Where Restaurant Food Poisoning Comes From

A ban is a judgement about where the danger lies. The outbreak records show plainly where it does lie, and it is worth setting the dog beside it.

### 5.1 Four outbreaks in ten

In the American report on 800 outbreaks, norovirus was the commonest cause where a cause was found, at 47%, and Salmonella was second. Investigators identified what had gone wrong in 500 of the outbreaks. In 205 of those, or 41%, the cause included a food worker who was ill or infectious and who touched the food. Norovirus lives in people and no other animal carries it. A review of the research published between 2003 and 2017 found 47 reports of norovirus outbreaks definitely caused by a food handler, in settings from weddings to military bases.

*The commonest cause of restaurant food poisoning walks in through the staff entrance.*

### 5.2 Hand washing

The same body of research explains why. Observers from the Centers for Disease Control and Prevention watched 321 food workers at their work and counted the moments when the food code required them to wash their hands, about nine times an hour. The workers tried to wash on 32% of those occasions and did it properly, with soap and drying, on 27%. When they were wearing gloves the figure fell to 16%.

This is a measure of proportion and not an attack on restaurant staff. The health code that will not let a dog lie on the floor of the dining room is, in the same building, complied with about a quarter of the time on the single practice that matters most.

*An inspector worried about the food would learn more from watching the hand basin than from looking under the tables.*

### 5.3 What a proportionate rule looks at

Modern food law is meant to be based on risk. It asks where contamination can happen and puts the control there. The European regulation follows that logic and stops at the kitchen door, because that is where the food is open.

A rule that follows the same logic into the dining room finds very little to control. The food is cooked and plated. The dog is on the floor. The one path between them is a person's hand, and a hand washing rule already covers it.

*Judged by the method food law uses everywhere else, the dog under the table does not make the list.*

---

## 6. The Real Reasons to Manage a Dog

None of this means a dog in a dining room needs no rules. It means the rules should be about the things that are real. There are three, and none of them is food poisoning.

### 6.1 Allergies

Dog allergen is real, and it is airborne. A British study of fourteen hospitals found it in the air of an outpatient waiting room on every day that samples were taken, in a building where dogs do not go. It comes in on the clothes of people who own one, and the study found it settled in the upholstered chairs. The same has been found in schools, cars and public transport.

That cuts both ways. A room that has never held a dog is not free of the allergen, but a room with a dog in it will hold far more, and a guest with asthma is entitled to know before sitting down. Our paper [Weighed and Found Wanting](https://rochstandard.com/research/the-hotel-dog-weight-limit/) sets out the allergen research in more detail. The answer is a sign on the door, a part of the room kept free of dogs, and hard seating where dogs are welcome.

*A guest with an allergy needs to be told and to be given somewhere else to sit, and a ban does neither better than a sign and a seating plan.*

### 6.2 Bites, trips and hot plates

Dog bites are common. American emergency departments treated about 334,000 a year in the 1990s, and hospital admissions for bites in England more than doubled between 1998 and 2018. No study counts bites in restaurants, and most bites happen at home, to people who know the dog. But a crowded room, strange dogs, food on the floor and children at eye level are the conditions in which a nervous dog snaps.

The plainer hazard is physical. A dog lying in an aisle is in the way of a server carrying hot plates, and a long lead across a walkway will catch somebody's foot.

*The injuries worth planning for are a bite and a fall, and both are prevented by where the dog is put and how short the lead is.*

### 6.3 The guest who did not bring a dog

The diner at the next table has a say. A guest who dislikes or fears dogs should not find out by sitting beside one, and researchers have begun to study how customers weigh the risks they see in restaurants that welcome pets. The owners themselves worry about behaviour and space, as the Toronto research behind [A Place at the Table](https://rochstandard.com/research/dining-with-dogs/) found.

What settles it is a clear rule, stated before the guest arrives and applied the same way to everyone. The Roch Dog Friendly Standard calls this the Friendly Dog Policy: a dog is welcome for as long as it behaves, whatever its size or breed, and the business keeps the right to ask any dog to leave.

*A room works for everyone when the rules are known in advance and the staff are willing to enforce them.*

### 6.4 The dog itself

The last interest is the dog's. A busy dining room is loud, crowded and full of food it may not have, and not every dog is comfortable in one. A dog that pants, paces or hides under the chair is telling its owner that it would sooner be somewhere else, and an anxious dog is the one most likely to bark or snap.

An operator can help with small things. A table at the edge of the room, away from the kitchen door and the main walkway, suits a dog far better than one in the middle. So does a bowl of water brought without being asked, and room to lie down flat. The owner remains the best judge of whether their dog can manage a meal out, and most of them judge it well.

*A calm dog in a quiet corner is safer and happier than a nervous one in the middle of the room, and the seating plan decides which you get.*

---

## 7. What the Countries That Changed Their Minds Did

The governments that have opened their dining rooms have each had to write down the conditions. Set side by side, their lists are nearly identical, and together they are the nearest thing there is to an agreed code of practice.

### 7.1 Outdoors first

Florida went first, in 2006, with a law letting any town or county exempt the outdoor areas of its restaurants from the animal rule. Other American states followed with patio laws of their own, California among them. Australia lifted its ban on dogs in outdoor dining areas nationally in 2012, on the strength of the risk assessment described in section 4. In 2022 the Food and Drug Administration wrote the same exception into the American model code, where the local authority approves it. Singapore removed the need for special approval to seat guests with pets in outdoor refreshment areas from 1 January 2025.

*For twenty years the countries with a ban have been taking it apart from the outside in.*

### 7.2 Then indoors

In 2026 two governments went further. South Korea changed its food hygiene rules from 1 March to let ordinary restaurants, snack bars and bakeries admit dogs and cats. Hong Kong, where bringing a dog onto food premises had been an offence punishable by a fine and imprisonment, changed its regulation in May so that restaurants can apply for permission to admit dogs. Its government said it had reviewed public health, public opinion and the views of the trade, and concluded that conditions now favoured allowing dogs "under appropriate regulation".

*Two of the strictest food regimes in Asia have looked at the evidence and decided that a dog may sit indoors.*

### 7.3 The conditions they set

The American model code asks an operator to think about the dog's vaccinations, to keep it restrained, to keep it away from staff, food, utensils, tables and chairs, to post a sign, and to keep anything served to a dog apart from what is served to people. South Korea requires proof of vaccination at the door, a sign at the entrance and space between tables. Hong Kong requires a sign at the entrance so that, in its words, the public can make an informed choice. Australia's assessment relied on dogs staying out of food preparation areas and on ordinary hygiene by staff.

None of these governments has yet published a study of what happened next, and that is the largest gap in the evidence. But none has reversed its decision either, and the first of them has had nearly twenty years to do so.

*The same five conditions recur everywhere: a sign, a lead, the floor, separate dishes and clean hands.*

### 7.4 The dogs that were always allowed

One group of dogs has been eating out legally all along. The 1962 American code that closed the dining room to animals opened it, in the same sentence, to guide dogs. Disability law has since widened that to assistance dogs of many kinds, in almost every country, and they sit in restaurants, cafes and hotel dining rooms every day.

Those dogs are well trained, but training does not change what a dog carries on its coat or its paws. If a dog on the floor of a dining room were a danger to the food, sixty years of assistance dogs would have shown it, and we found no outbreak attributed to one. The law has always accepted that a dog can lie under a table safely. It has only been particular about which dog.

*Assistance dogs have been the dining room trial that nobody set out to run, and it has lasted sixty years.*

---

## 8. What to Do in Your Dining Room

What follows is our reading of the evidence for an operator. It is advice on good practice and not on the law, which differs from place to place and which you should check with your local authority before changing anything.

### 8.1 Know which rule you are under

In Britain and across most of Europe the decision is yours, and the legal duty is to keep the dog out of the kitchen and food stores. In the United States, Canada and Australia a dog is generally lawful outdoors only, and often only with local approval. Assistance dogs are a separate matter and must be admitted almost everywhere.

If the law leaves you the choice, the evidence gives you no health reason to refuse. If it does not, the terrace, the garden, the lobby and the hotel bar's outdoor tables are where your welcome begins, and your local authority is the place to ask about the rest.

*Find out exactly what your own law forbids. It is often less than your staff have been told.*

### 8.2 The five rules that do the work

The dog stays on the floor, on a short lead, beside its owner's chair and out of the aisle. It does not sit on a chair or a bench, and it does not eat from a plate or drink from a glass.

Anything the dog is given comes in its own bowl. That bowl is kept and washed apart from the tableware, or is disposable.

Staff who touch a dog wash their hands before they touch anything else. This is the one path from the dog to the food, and it is closed by a rule your kitchen already has.

A notice at the door and on the booking page says that dogs are welcome, so that a guest with an allergy or a fear of dogs can choose. Part of the room is kept free of dogs.

Any dog that barks, lunges or will not settle is asked to leave, whatever its size or breed, and the staff know they have your backing to say so.

*These five rules answer every real risk in this paper, and none of them costs more than a sign and a set of bowls.*

### 8.3 If you feed the dog

A menu for dogs is good business, as [The Highest Margin Plate in the Hotel](https://rochstandard.com/research/the-hotel-dog-food-menu/) shows. The raw feeding research in section 3 adds one firm rule to it: serve dogs cooked food only. Prepare it with the same care as any other dish, keep the bowls separate from the plates, and you have removed the one hazard the science has actually demonstrated.

*Cook what you serve to a dog, and serve it in the dog's own bowl.*

### 8.4 For hotels

A hotel has more rooms in play than a restaurant, and its guest cannot go home to eat. A guest who has paid to bring a dog and is then barred from breakfast, the bar and the lounge has been given a room and little else. Travellers notice this. The research behind [More Than Permission](https://rochstandard.com/research/what-pet-friendly-means/) found that in guests' eyes a welcome confined to the terrace fares little better than an outright restriction.

Open at least one indoor room where a guest can eat and drink with the dog beside them. The lounge and the bar are the easy ones. Give the breakfast room a section for guests with dogs, and say so in the dog policy, which [The Fine Print](https://rochstandard.com/research/the-hotel-pet-policy/) explains how to write.

*The guest with a dog has to eat somewhere. Make it a table in your hotel.*

---

## 9. What Nobody Has Measured

We have tried throughout to say where the evidence stops, and this section gathers the limits in one place.

### 9.1 No dining room has been tested

No published study has compared the surfaces, the air or the illness records of dining rooms that admit dogs with those of dining rooms that do not. No study has followed inspection results before and after a patio law. The paw and shoe studies are small, and they measured marker bacteria, not viruses or parasites. The outbreak records cannot see a single case of illness that nobody reported.

So the position is plain. The ban was adopted without evidence, sixty years of looking have produced none, and every piece of indirect evidence points the same way. The one thing missing is a direct test, and the people with the power to run it are the health authorities who enforce the rule.

*The burden of proof belongs to the rule, and in sixty years the rule has not met it.*

### 9.2 The studies that would settle it

Three pieces of work would close the gap. The first is a comparison of matched cafes or pubs, with and without dogs, swabbing tables, floors, staff hands and the air during service. Britain or Germany could run it tomorrow. The second is a follow up of the Hong Kong and South Korean schemes, comparing inspection results and complaints in the restaurants that admit dogs with those that do not. The third is a measurement of dog allergen in the air of a working dining room, to show how far a section kept free of dogs protects the guest who needs it.

We would gladly work with any university or health authority that wants to carry out one of them.

*Each of these studies is cheap and short, and none of them has been done.*

### 9.3 What we did not read

Our history rests on the documents we could obtain. We read the American codes of 1940, 1943 and 1962 and the current Food Code, the British regulations of 1970 and 1995, and the European texts. We did not read the American drafts of 1935 and 1938 or the edition of 1976, the British regulations of 1955 and 1960, or the original 1969 text of the international code, none of which we could find in full. The appendix lists what we read and how.

*Where we have not read a document ourselves, we have not relied on it.*

---

## 10. Conclusion

I draw four conclusions from this work.

The first is that the dining room ban was never a scientific finding. In the United States, where it began, the rule stopped at the kitchen for a quarter of a century and was widened in 1962 by a drafting committee that gave one general sentence as its reason. Britain never adopted it. The international codes treat animals as a matter of pest control in factories. A rule with that history is a custom, and it deserves the respect due to a custom and no more.

The second is that the evidence gathered since has not rescued it. Dogs carry germs, and raw fed dogs carry more, and that justifies keeping every dog away from food preparation and off the tableware. But the outbreak records are silent on dogs in dining rooms, the one regulator to assess the question found the risk very low to negligible, and the studies that compare dogs with people keep finding the people dirtier.

The third is that the food code is looking in the wrong place. Sick staff and unwashed hands cause a large share of restaurant food poisoning, and the hand washing rule is followed about a quarter of the time. A health authority with one hour to spend in a restaurant would protect more diners at the hand basin than at the door.

The fourth is that the real risks are manageable and already have an agreed answer. Allergies, bites, trips and the comfort of other guests are dealt with by a sign, a short lead, a place on the floor, separate dishes and clean hands. Every government that has opened its dining rooms has arrived at the same short list.

So I have two requests. To anyone who runs a dining room where the law allows a choice: let the dog in, under those five rules, and tell your guests you have done it. To the health authorities that still forbid it: do what Western Australia did, and ask whether the rule is based on risk. If it is, the evidence should be published. If it is not, the dog should be allowed to lie under the table.

---

## Appendix: How We Did This

**The history.** We read the model codes in the scanned originals held by the Internet Archive: the Ordinance and Code Regulating Eating and Drinking Establishments in its 1940 and 1943 editions, and the Food Service Sanitation Manual of 1962. The current rule and its stated reason are quoted from the Food and Drug Administration's Food Code 2022 and its Annex 3. The British regulations of 1970 were read in the Wellcome Collection's scan, and those of 1990 and 1995, the European directive of 1993 and Regulation 852/2004 on legislation.gov.uk. The international code is quoted from its 1997 revision as reproduced by the Food and Agriculture Organization.

**The studies.** We searched PubMed and Crossref with 45 queries covering what dogs carry, raw feeding, dogs in hospitals and other shared spaces, outbreaks from animal contact, food workers and hand washing, shoes, allergens, bites and customer attitudes. Where a study is open access we read it in full. Where it is not, we read the published abstract and say no more than the abstract supports. The figures for the Guelph bowl experiment are taken from the review by Davies and colleagues, which reports them.

**The outbreak records.** The three reports from the Centers for Disease Control and Prevention were read in full, and the two restaurant outbreak reports were searched for the words dog, pet and animal.

**Our own figures.** The shares of hotels welcoming dogs in the bar, restaurant and lounge come from the Roch Dog directory as it stood on 18 November 2026. Each hotel's answer is what the hotel told us when it was assessed. A bar or restaurant whose own outdoor tables are the only part open to dogs is counted as open, so the North American figures overstate indoor access.

**The law.** Statements of the law are a summary for general readers and are not legal advice. The rules in South Korea and Singapore are taken from government announcements and press reports of them, and those in Hong Kong from the Food and Environmental Hygiene Department. The position in individual American states and Canadian provinces varies and should be checked locally.

---

## Sources

### Food Codes and Law
1. [United States Public Health Service (1940). Ordinance and Code Regulating Eating and Drinking Establishments](https://archive.org/details/micro_IA41159428_0693)
2. [United States Public Health Service (1943). Ordinance and Code Regulating Eating and Drinking Establishments, 1943 Recommendations](https://archive.org/details/micro_IA41159428_0067)
3. [United States Public Health Service (1962). Food Service Sanitation Manual. Public Health Service Publication No. 934](https://archive.org/details/micro_IA41159442_0100)
4. [Food and Drug Administration (2022). Food Code 2022, section 6-501.115 and Annex 3](https://www.fda.gov/media/164194/download)
5. [The Food Hygiene (General) Regulations 1970, S.I. 1970/1172](https://archive.org/details/b32185236)
6. [The Food Safety (General Food Hygiene) Regulations 1995, S.I. 1995/1763](https://www.legislation.gov.uk/uksi/1995/1763/schedules/made)
7. [Regulation (EC) No 852/2004 on the hygiene of foodstuffs, Annex II](https://www.legislation.gov.uk/eur/2004/852/annexes)
8. [Codex Alimentarius. General Principles of Food Hygiene, 1997 revision](https://www.fao.org/4/w8088e/w8088e04.htm)
9. [Food Standards Australia New Zealand (2012). Proposal P1018: Companion Dogs in Outdoor Dining Areas, approval report](https://www.foodstandards.gov.au/sites/default/files/food-standards-code/proposals/Documents/P1018%20Companion%20Dogs%20in%20Outdoor%20Dining%20Areas%20AppR%20AMENDED%20150812.pdf)
10. [Florida Statutes, section 509.233](http://www.leg.state.fl.us/statutes/index.cfm?App_mode=Display_Statute&URL=0500-0599/0509/Sections/0509.233.html)
11. [Food and Environmental Hygiene Department, Hong Kong. Dogs in restaurants: background and legislation](https://www.fehd.gov.hk/english/licensing/dog_restaurants/legislative_framework.html)
12. [Singapore Food Agency. No additional approvals needed for patrons to dine with pets at outdoor refreshment areas from 1 January 2025](https://sfa.gov.sg/news-publications/newsroom/no-additional-sfa-approvals-needed-for-patrons-to-dine-in-with-pets-at-outdoor-refreshment-areas-from-1-jan-2025)
13. [Asia News Network (2026). Pet-friendly dining set to expand in South Korea as regulations ease from March](https://asianews.network/pet-friendly-dining-set-to-expand-in-south-korea-as-regulations-ease-from-march/)

### What Dogs Carry
14. [Reimschuessel, R. and colleagues (2017). Multilaboratory survey to evaluate Salmonella prevalence in diarrheic and nondiarrheic dogs and cats in the United States between 2012 and 2014. Journal of Clinical Microbiology 55(5): 1350-1368](https://doi.org/10.1128/JCM.02137-16)
15. [Lefebvre, S. L. and colleagues (2006). Prevalence of zoonotic agents in dogs visiting hospitalized people in Ontario. Journal of Hospital Infection 62(4): 458-466](https://doi.org/10.1016/j.jhin.2005.09.025)
16. [Pintar, K. D. M. and colleagues (2015). A systematic review and meta-analysis of the Campylobacter spp. prevalence and concentration in household pets and petting zoo animals. PLOS ONE 10(12): e0144976](https://doi.org/10.1371/journal.pone.0144976)
17. [Runesvärd, E., Wikström, C., Fernström, L.-L. and Hansson, I. (2020). Presence of pathogenic bacteria in faeces from dogs fed raw meat-based diets or dry kibble. Veterinary Record 187(9): e71](https://doi.org/10.1136/vr.105644)
18. [Bernaquez, I. and colleagues (2025). Dogs fed raw meat-based diets are vectors of drug-resistant Salmonella infection in humans. Communications Medicine 5: 214](https://doi.org/10.1038/s43856-025-00919-2)
19. [Weese, J. S. and Rousseau, J. (2006). Survival of Salmonella Copenhagen in food bowls following contamination with experimentally inoculated raw meat. Canadian Veterinary Journal 47(9): 887-889](https://pmc.ncbi.nlm.nih.gov/articles/PMC1555674/)
20. [Davies, R. H., Lawes, J. R. and Wales, A. D. (2019). Raw diets for dogs and cats: a review, with particular reference to microbiological hazards. Journal of Small Animal Practice 60(6): 329-339](https://doi.org/10.1111/jsap.13000)

### Dogs in Shared Spaces
21. [Eisenstein, T. and colleagues (2025). Enteric disease outbreaks associated with animal contact, United States, 2009-2021. MMWR Surveillance Summaries 74(3): 1-12](https://doi.org/10.15585/mmwr.ss7403a1)
22. [Vos, S. J., Wijnker, J. J. and Overgaauw, P. A. M. (2021). A pilot study on the contamination of assistance dogs' paws and their users' shoe soles. International Journal of Environmental Research and Public Health 18(2): 513](https://doi.org/10.3390/ijerph18020513)
23. [Janezic, S., Mlakar, S. and Rupnik, M. (2018). Dissemination of Clostridium difficile spores between environment and households: dog paws and shoes. Zoonoses and Public Health 65(6): 669-674](https://doi.org/10.1111/zph.12475)
24. [Gutzeit, A. and colleagues (2019). Would it be safe to have a dog in the MRI scanner before your own examination? European Radiology 29(2): 527-534](https://doi.org/10.1007/s00330-018-5648-z)
25. [Lefebvre, S. L., Reid-Smith, R. J., Waltner-Toews, D. and Weese, J. S. (2009). Incidence of acquisition of methicillin-resistant Staphylococcus aureus, Clostridium difficile, and other health-care-associated pathogens by dogs that participate in animal-assisted interventions. Journal of the American Veterinary Medical Association 234(11): 1404-1417](https://doi.org/10.2460/javma.234.11.1404)
26. [Murthy, R. and colleagues (2015). Animals in healthcare facilities: recommendations to minimize potential risks. Infection Control and Hospital Epidemiology 36(5): 495-516](https://doi.org/10.1017/ice.2015.15)

### Food Workers and Restaurant Outbreaks
27. [Moritz, E. D. and colleagues (2023). Foodborne illness outbreaks at retail food establishments, National Environmental Assessment Reporting System, 2017-2019. MMWR Surveillance Summaries 72(6): 1-11](https://doi.org/10.15585/mmwr.ss7206a1)
28. [Holst, M. M. and colleagues (2025). Contributing factors of foodborne illness outbreaks, National Outbreak Reporting System, United States, 2014-2022. MMWR Surveillance Summaries 74(1): 1-12](https://doi.org/10.15585/mmwr.ss7401a1)
29. [Green, L. R. and colleagues (2006). Food worker hand washing practices: an observation study. Journal of Food Protection 69(10): 2417-2423](https://doi.org/10.4315/0362-028X-69.10.2417)
30. [Hardstaff, J. L. and colleagues (2018). Foodborne and food-handler norovirus outbreaks: a systematic review. Foodborne Pathogens and Disease 15(10): 589-597](https://doi.org/10.1089/fpd.2018.2452)

### Allergens, Bites and Diners
31. [Custovic, A. and colleagues (1998). Domestic allergens in public places III: house dust mite, cat, dog and cockroach allergens in British hospitals. Clinical and Experimental Allergy 28(1): 53-59](https://doi.org/10.1046/j.1365-2222.1998.00183.x)
32. [Weiss, H. B., Friedman, D. I. and Coben, J. H. (1998). Incidence of dog bite injuries treated in emergency departments. JAMA 279(1): 51-53](https://doi.org/10.1001/jama.279.1.51)
33. [Tulloch, J. S. P. and colleagues (2021). English hospital episode data analysis (1998-2018) reveal that the rise in dog bite hospital admissions is driven by adult cases. Scientific Reports 11: 1767](https://doi.org/10.1038/s41598-021-81527-7)
34. [Zhang, X., Shao, X. and Zheng, Y. (2027). Dining with dogs or doubts? Understanding customers' perceptions of pet-friendly restaurants through perceived risk and protection motivation theory. International Journal of Hospitality Management 140: 104819](https://doi.org/10.1016/j.ijhm.2026.104819)

### Roch Dog Research
35. [Roch Dog Friendly Standard (RDFS-02)](https://rochstandard.com/standard/)
36. [RDWP-DWD-01: A Place at the Table](https://rochstandard.com/research/dining-with-dogs/)
37. [RDWP-FD-01: The Highest Margin Plate in the Hotel](https://rochstandard.com/research/the-hotel-dog-food-menu/)
38. [RDWP-PFD-01: More Than Permission](https://rochstandard.com/research/what-pet-friendly-means/)
39. [RDWP-WL-01: Weighed and Found Wanting](https://rochstandard.com/research/the-hotel-dog-weight-limit/)
40. [RDWP-HPP-01: The Fine Print](https://rochstandard.com/research/the-hotel-pet-policy/)

---

## Published by Roch Dog

Roch Dog is the independent ratings body for dog friendly hotels. It assesses hotels against the Roch Dog Friendly Standard and publishes its research free at rochstandard.com/research. This paper is published under a Creative Commons Attribution licence (CC BY 4.0). Research enquiries: research@rochdog.com.
