Almost every page answering this question gives you a list of ten breeds and a warm feeling. This one starts somewhere else, because the research points somewhere else.
The breed you choose is close to the least important decision you will make here. What matters more is how much dog you are taking on, what you can realistically manage on a bad week, and whether you are getting a dog for reasons that will still hold in six months.
The breeds are further down, and they are answered properly. But you deserve the honest version first.
I am not a clinician and none of this is medical advice. It is what the published evidence says, linked so you can read it yourself.
What the evidence actually shows
This is where most articles wave a hand at “studies show.” Here are the studies.
The largest synthesis of this question is **Brooks and colleagues, BMC Psychiatry, 2018** — a systematic review of the evidence on companion animals and people living with mental health problems.
Their conclusion is genuinely positive: “pets provide benefits to those with mental health conditions.” Fifteen of the seventeen included studies reported positive aspects.
And then the parts nobody quotes.
There were no randomized controlled trials. In the reviewers’ own words, “There were no RCTs evaluating the impact of pet ownership on diagnosed mental health conditions.” That is not a small gap. It means the strongest form of evidence for whether a thing works does not exist here.
The quantitative results were mixed — studies showed positive, negative and neutral impacts, and the qualitative studies averaged a quality rating of 5.5 out of 10.
And nine of the seventeen studies reported negative elements, specifically “the practical and emotional burden of pet ownership and the psychological impact that losing a pet has.”
So the honest summary is this: people with mental health problems consistently describe their animals as important to them, the research supports that, and the evidence that acquiring an animal improves a diagnosed condition does not really exist yet.
The finding that changes the question
There is a more recent piece of work worth knowing about, and it is uncomfortable in a useful way.
**Zablan, Melvin and Hayley, BMC Psychology, 2024** compared dog owners with owners of other companion animals, looking at physical activity, loneliness, depression and anxiety.
Dog owners were more active. No surprise there. But the detail matters:
**Only the physical activity people did independently of the dog was associated with better mental health outcomes.**
Not the dog walking. The other activity. The authors’ reading is that people who choose dogs tend to be people who live more active lives generally — and it is that broader activity, not the dog specifically, that tracks with feeling better.
And they say the practical implication out loud:
“a person acquiring a dog with the sole purpose of improving their mental health is unlikely to experience mental health benefits.”
That is the authors of a pro-companion-animal study telling you that a dog acquired as a treatment probably will not work as one.
It is a cross-sectional study and cannot prove causation — they say so themselves. But it is the most direct answer available to the question you actually came here with.
So should you not get a dog?
That is not what any of this says, and it would be a poor reading of it.
What the evidence undercuts is a specific claim: that acquiring a dog is an intervention for depression. It is not. It has never been shown to be, there are no trials showing it, and the people selling you that idea have not read the literature.
What a dog genuinely does is more ordinary and more durable than that.
It puts a shape on the day. Something has to be fed at a particular time whether or not you feel like getting up. There is a reason to open the front door. There is a living thing in the room that is pleased you exist, on days when that is in short supply.
Those are real. They are also not a treatment, and the difference is not pedantry — it decides whether you are disappointed in six months.
A dog can make a hard day more liveable. It cannot make a hard year into an easy one, and it is not a substitute for care that treats the thing you actually have.
If you are struggling badly right now, a dog is not the first call. Talk to your doctor or a mental health professional. That is not a brush-off — it is the same advice I would want given to me, and the two are not in competition. Plenty of people do both.
The risk nobody writes about
Here is the part the breed lists leave out, and it is the reason this page exists.
Getting a dog while you are depressed can go badly.
A dog is a decade or more of daily obligation. It needs walking on the days you cannot get out of bed. It needs feeding when you are not feeding yourself properly. It costs money every month and a great deal of money without warning. It will get ill. It will die, and Brooks and colleagues specifically identified the psychological impact of losing a pet among the negatives their review found.
And there is a specific trap in depression. When you cannot meet the dog’s needs, the failure does not stay neutral. It becomes evidence — one more thing you are getting wrong, attached to a living creature who deserved better. A dog acquired to lift your mood can become a source of guilt that deepens it.
I am not saying do not get a dog. I am saying make the decision the way you would make it if you were well, and ask three questions honestly:
- On my worst week, who walks and feeds this dog? If the honest answer is nobody, that needs solving before the dog arrives, not after.
- Can I absorb an unexpected veterinary bill? The routine costs are predictable. The emergency is not, and the money question is set out in how much does a service dog cost — much of which applies to any dog.
- Would I still want this animal if my mood improved next month? If the honest answer is no, that is important information.
If those answers are good, the rest of this page is useful to you. If they are not, the most caring thing you can do for a dog that does not exist yet is to wait.
What an emotional support animal actually gets you
This matters because the old version of this page got it wrong, and so do most pages on the subject.
| Emotional support animal | |
|---|---|
| Housing | Yes. A reasonable accommodation under the Fair Housing Act, and HUD states plainly that “breed, size, and weight limitations may not be applied to an assistance animal” |
| Flights | No. Removed from the Air Carrier Access Act on 11 January 2021. An ESA now flies as a pet, under pet rules and pet fees |
| Shops, restaurants, public places | No. The ADA covers task-trained service dogs. An ESA has no public access right |
| Training required | None. That is the definition — an ESA is not trained to do work or perform tasks |
The flying change is the one that catches people. If you have read that your ESA can fly with you, you have read something that stopped being true over five years ago, and we go through what replaced it in flying with a large service dog.
The housing right is real and it is worth knowing properly — can a landlord refuse a service dog covers it, emotional support animals: requirements and benefits covers the rest, and what disabilities qualify for an emotional support animal covers who is eligible in the first place.
Name the incentive
Depression is the single most valuable search term in the emotional support animal business, and there is an entire industry selling letters to the people typing it.
An ESA letter is a document from a licensed mental health professional who is treating you, confirming a disability-related need. What is sold online is frequently a form filled in after a five-minute questionnaire by someone who will never see you again.
A letter from a clinician who knows you is a clinical judgment. A letter from a website is a receipt.
The tell is simple. A real assessment can say no. If the process cannot produce a “no,” it is not assessing anything — and that is the same test we apply to registries in how to identify a fake service dog.
Nothing on this site is paid for by a letter service, and nothing ever will be. The temptation is largest on exactly this page, which is why it is worth saying here.
Now: the breed question
If you have got this far and the answers were good, here is how to choose — and it is not a top ten.
Match the dog’s demand to your worst week, not your best day.
That single sentence does more work than any breed list. The failure mode is not choosing the wrong breed; it is choosing a dog whose needs you can meet when you are well and cannot meet when you are not.
| What to weigh | Why it decides things |
|---|---|
| Daily exercise need | A dog needing two hours is a commitment on the days you cannot manage twenty minutes |
| Grooming demand | Coat care is a recurring cost in money, time and attention |
| Separation tolerance | A dog that cannot be left alone narrows your life rather than widening it |
| Size and strength | Relevant to handling, housing, and what you can physically manage on a bad day |
| Age at acquisition | An adult dog is a known quantity. A puppy is eighteen months of work |
That last row deserves more attention than it gets. Adopting an adult dog with a known temperament removes most of the risk this page is about. You can meet the animal, see how it actually is, and skip the phase that costs the most sleep.
The specific mismatch to avoid: the breeds most often recommended for depression are frequently the high-energy ones, on the reasoning that they will get you out of the house. That reasoning inverts the problem. A dog that needs two hours of exercise is not a solution to low motivation — it is a daily test of it, and it is the profile most likely to end in guilt.
For the breeds themselves, worked through honestly by demand level rather than by charm, we have a much fuller treatment: 64 best emotional support dog breeds, grouped by how much dog you are actually taking on — including a section on the flat-faced breeds that anyone considering one should read first.
If tasks would help more than company
One more thing worth knowing, because the categories get blurred and the difference is large.
**If what you need is a dog that does something specific — interrupts a spiral, gets you medication, wakes you, gets you out of a situation — that is task work, and it is a psychiatric service dog rather than an emotional support animal.** Different legal category, different training, different rights.
The line is set out in emotional support animal vs service dog, the tasks themselves in what PTSD service dogs do, and whether you would qualify in do I qualify for a service dog.
Frequently asked questions
What is the best emotional support dog breed for depression?
There is no best breed, and the research does not support naming one. What predicts whether it goes well is matching the dog’s daily demands to what you can manage on a bad week, not on a good day. An adult dog with a known temperament is usually a safer choice than a puppy, and a lower-demand dog is usually safer than a high-energy one, because a dog needing two hours of exercise tests low motivation rather than solving it.
Do emotional support animals actually help with depression?
The evidence is more mixed than most pages admit. A 2018 systematic review in BMC Psychiatry found pets provide benefits to people with mental health conditions, but also that there were no randomized controlled trials on diagnosed conditions, that quantitative results were mixed, and that nine of seventeen studies reported negative aspects including the burden of care and the impact of losing a pet.
Will getting a dog cure my depression?
No, and it is worth being clear about that. A dog is not a treatment and has never been shown to be one. The authors of a 2024 BMC Psychology study put it directly: someone acquiring a dog with the sole purpose of improving their mental health is unlikely to experience mental health benefits. A dog can make a hard day more liveable, which is real and worth having, but it is not a substitute for treatment.
Can my emotional support dog fly with me?
No. The Department of Transportation removed emotional support animals from the Air Carrier Access Act in a rule that took effect on 11 January 2021. Since then an ESA flies as a pet, under the airline’s pet policy and pet fees. Only dogs individually trained to do work or perform tasks are covered as service animals on flights.
Can a landlord refuse my emotional support animal?
Generally no. An assistance animal in housing is treated as a reasonable accommodation under the Fair Housing Act rather than as a pet, and HUD guidance states that breed, size and weight limitations may not be applied to an assistance animal. That is the one substantial legal right an ESA carries, and it is worth understanding properly before you need it.
Do I need an ESA letter, and where should I get one?
A letter comes from a licensed mental health professional who is actually treating you and can speak to a disability-related need. Letters sold online after a brief questionnaire are a different product entirely. A useful test is whether the process is capable of saying no. If it cannot, it is not assessing anything, and a document that cannot be refused is not evidence of anything either.
Is a puppy or an adult dog better if I am depressed?
An adult dog is usually the safer choice. Its temperament is already visible rather than a prediction, you can meet it before committing, and you skip roughly eighteen months of the most demanding work there is. Puppies are hardest in exactly the ways depression makes life hardest already, which is broken sleep, constant supervision and slow progress.
What if I cannot look after the dog on a bad day?
That is the question to answer before the dog arrives rather than after. Having someone who can step in, whether a partner, family member, neighbor or paid walker, turns the single biggest risk into a manageable one. The failure mode this page warns about is not a bad dog, it is an unmet need becoming one more thing to feel guilty about.
Should I get a psychiatric service dog instead?
That depends on whether you need company or trained help. An emotional support animal helps by being present and requires no training. A psychiatric service dog is trained to perform specific tasks, such as interrupting an episode or retrieving medication, and carries public access rights an ESA does not. They are different categories with different costs and different obligations.
The short version
No breed is best. Match the dog’s demands to your worst week, choose an adult over a puppy where you can, and be wary of high-energy breeds recommended precisely because they will force you out.
The evidence is warmer than it is strong. People with mental health problems consistently value their animals, and there are no trials showing that getting one improves a diagnosed condition. The most direct study available says a dog acquired purely to improve mental health probably will not.
Ask the three questions honestly — who covers the bad weeks, can you absorb an emergency bill, and would you still want this dog if you felt better next month.
And know what you are actually getting. Housing rights, yes. Flights, not since January 2021. Public access, no.
A dog is not medicine. It is a decade of company, obligation and expense — and for the right person, at the right time, entirely worth it.
Sources: Brooks HL et al., The power of support from companion animals for people living with mental health problems: a systematic review and narrative synthesis of the evidence, BMC Psychiatry 2018;18:31 · Zablan K, Melvin G, Hayley A, Dog ownership, physical activity, loneliness and mental health, BMC Psychology 2024 · HUD FHEO assistance animals notice · DOT Final Service Animal Rule, 85 FR 79774.