Hospital Hours: Monday-Friday: 8am-6pm | Saturday: 8am-4pm
(519) 896-0532
There is no single best anxiety medication for dogs. There's a best one for your dog, and the pattern of the anxiety decides that more than the drug does. Anxiety that runs every day needs a daily medication that takes weeks to build up. Anxiety that arrives with the fireworks needs something you give two hours before and that wears off by morning. Get that split right and most of the rest follows.
Key Takeaways
Don't give your dog human anxiety medication from your own cupboard. Human antidepressants and benzodiazepines are a common cause of poisoning in dogs, and the dose that helps a person can cause tremors, seizures or dangerously high body temperature in a dog. If your dog has already swallowed something, call us at (519) 896-0532 or use our emergency page after hours. Bring the bottle.
Start by answering one question: is your dog anxious most days, or only when something specific happens?
Everything below follows from that answer.
| Medication | Best suited to | How fast it works | Dosing | Main caution |
|---|---|---|---|---|
| Fluoxetine (Prozac, Reconcile) | Daily, ongoing anxiety. The best evidence is in separation anxiety | 4 to 8 weeks | Once daily, every day | Reduced appetite and lethargy are common in the first few weeks1 |
| Clomipramine (Clomicalm) | Separation anxiety and compulsive behaviours | 4 to 8 weeks | Twice daily, every day | Not for dogs with heart rhythm disease or glaucoma11 |
| Trazodone | Situational fear. Vet visits, travel, storms, crate rest after surgery | 1 to 2 hours | Before the trigger, or twice daily | Sedation. Needs care alongside an SSRI11 |
| Gabapentin | Situational fear, and especially useful when pain is also in the picture | 1 to 2 hours | Before the trigger, or two to three times daily | Wobbliness and sedation, more likely above 30 mg/kg8 |
| Alprazolam (a benzodiazepine) | Short, predictable episodes of panic. Fireworks night | 30 to 60 minutes | Before the trigger, repeated if needed | Can release inhibitions and worsen aggression in some dogs11 |
These get combined all the time. A dog on daily fluoxetine for separation anxiety may still get trazodone on the morning of a vet visit. That's a normal plan and often the one that works.
The daily medications are slow on purpose. If you start fluoxetine on Monday and stop it on Friday because nothing has changed, you've learned nothing about whether it works. Stopping early is the most common reason a good medication gets abandoned.
Anxiety is the anticipation of something bad. Fear is the response to something that's actually happening. In practice the two blur, and the medications overlap.
What owners usually see is pacing, panting, drooling, trembling, barking or howling with nothing there, destruction aimed at doors and windows, house soiling in a housetrained dog, or a dog who won't settle and won't eat.12 Some dogs go the other way and shut down completely, which is easy to miss and often read as good behaviour.
Before I reach for a prescription I want to rule out pain. A dog who is sore is a dog who can't settle, and restlessness from arthritis looks a great deal like restlessness from anxiety. That's also why gabapentin sometimes works better than its anxiety data alone would predict, since it treats the pain at the same time.
If you want help telling normal stress from a problem worth treating, our guide to why dogs pant and pace goes through what to look for.
Separation anxiety is panic at being left alone. The giveaway is timing. The destruction happens in the first thirty minutes after you leave, it's aimed at exit points rather than random objects, and the same dog is fine when a person is home.
This is the one type of canine anxiety with FDA approved drugs behind it, and it's also the one that most clearly needs a behaviour plan running alongside.
Thunderstorms, fireworks, construction noise, etc. This is the most common anxiety problem in dogs by some distance. In a survey of 13,700 Finnish pet dogs, noise sensitivity had a prevalence of 32%, ahead of general fearfulness and every other anxiety trait measured.10
Noise fear also tends to get worse with age rather than better, so a dog who was mildly bothered at three can be genuinely panicked at eight.
Fluoxetine is a selective serotonin reuptake inhibitor. It blocks the reabsorption of serotonin at the synapse, which leaves more of it available, and over several weeks the brain adjusts. That adjustment period is why it takes so long to work.
It's given once daily and I'd allow four to eight weeks before judging it.
The approval study enrolled 229 dogs across sites in the United States and Canada, 117 on fluoxetine and 112 on a control tablet, and every dog in both arms got behaviour modification. The proportion of dogs whose overall separation anxiety score improved was statistically higher on fluoxetine than on the control tablet in seven of the eight weeks.12
Reduced appetite and flatness are common in the first few weeks, and they're the reason some owners stop. There are numbers on that further down.
Clomipramine is a tricyclic antidepressant. It works on both serotonin and noradrenaline, and it was the first drug approved in North America for canine separation anxiety.4 It's given twice daily.
In the trial that supported it, 95 dogs were randomised to standard dose clomipramine, low dose clomipramine or placebo for two to three months, and all of them received behavioural therapy. Dogs on the standard dose improved at least three times faster than placebo for destruction, house soiling and inappropriate urination.3
I'd avoid it in dogs with known heart rhythm disease or glaucoma, and it shouldn't be combined with an MAOI such as selegiline.11
An earlier version of this article said trazodone was less effective than fluoxetine or clomipramine and used less often. Both halves of that were wrong.
Trazodone does a different job from the daily medications. It works within one to two hours instead of weeks,11 which makes it the drug of choice for anxiety you can see coming: a vet appointment, a car journey, a thunderstorm forecast, or eight weeks of crate rest after orthopaedic surgery.
It's also, by a wide margin, the most prescribed of the three. In a review of 32.4 million canine records from US primary care hospitals between 2010 and 2020, trazodone was prescribed to 1.33% of all dogs, against 0.03% for clomipramine and 0.02% for fluoxetine. Among dogs carrying a behaviour problem label, 8.4% received trazodone.7
The evidence has caught up too. In a 2026 placebo controlled trial using a road travel model, a single 10 mg/kg dose of trazodone given two hours before travel lowered serum cortisol, lip licking, global anxiety scores, urination, yawning and restlessness compared with placebo.5
We have a fuller write-up of trazodone for dogs if you want the detail.
Alprazolam is the one I reach for most in this class. Benzodiazepines work fast, usually within thirty to sixty minutes, and they wear off quickly, which suits a short and predictable event like fireworks night.
Two things make them harder to use than they look. Benzodiazepines can release behaviour a dog would normally hold back, and in some dogs that means excitement or aggression rather than calm.11 A dog who is fearful and already prone to snapping is not an automatic candidate.
They also don't erase the memory of a frightening event, so they work best given before the trigger rather than after the panic has started.
Increased appetite is a common and fairly harmless side effect. Ataxia and mild sedation are the others.11
Gabapentin is often described as working through GABA. It doesn't.
Despite the name, gabapentin has no meaningful activity at GABA receptors. It binds the α2δ subunit of voltage gated calcium channels, which reduces the release of excitatory neurotransmitters from the nerve ending.13 That's also why it works for nerve pain: it's the same mechanism.
For anxiety the evidence is mixed.
A double blind placebo controlled crossover trial in 18 dogs with storm phobia found a real reduction in fear responses on gabapentin at 25 to 30 mg/kg given at least 90 minutes before the storm, with ataxia the most frequent side effect.6 A separate survey of 50 owners whose dogs were prescribed gabapentin at a behaviour practice found 72% rated it moderately or very effective, though 70% also reported a side effect, most often sedation.8
Against that, the 2026 travel study that showed a clear benefit from trazodone found nothing from gabapentin at 30 mg/kg. On every physiological and behavioural measure it looked like placebo.5
My reading is that gabapentin is a reasonable choice for storm and noise fear, and a particularly good one when a painful older dog is part of the picture, but it isn't reliable enough to be a first pick for travel. It's also the one we use most often in cats.
Sileo is the only medication approved specifically for noise aversion in dogs, and it rarely comes up in articles like this one.
It's dexmedetomidine in an oromucosal gel, which means it goes between the cheek and the gum rather than being swallowed. It calms without knocking a dog out, and you can repeat it up to five times with at least two hours between doses.9
The approval study randomised 187 client owned dogs aged 2 to 17 years, with 144 evaluated for effectiveness, and used New Year's Eve fireworks as the trigger. The most common side effect was vomiting, in 4 dogs on the gel against 1 on the placebo gel.9
It was tested against fireworks, not thunderstorms, and the approval documentation says so directly.9 Plenty of dogs are fearful of both, and it's used for both, but the evidence behind it is a fireworks study.
These numbers come from the fluoxetine approval study, comparing 117 dogs on the drug with 112 on an identical tablet without it.1
| Side effect | On fluoxetine | On control tablet |
|---|---|---|
| Calm, lethargy or flatness | 45.3% | 17.0% |
| Reduced appetite | 29.1% | 10.7% |
| Shaking or tremor | 16.2% | 3.6% |
| Vomiting | 14.5% | 8.9% |
| Restlessness or hyperactivity | 13.7% | 6.3% |
The comparison matters more than the raw percentages. Vomiting happened in both groups at a similar rate, so most of it wasn't the drug. Lethargy and appetite loss clearly were.
Three interactions I want owners to know about:
There's direct evidence for this, and it's sitting in the fluoxetine approval file.
Alongside the study described above, the sponsor ran a second field study testing fluoxetine for separation anxiety without a behaviour modification plan. The FDA summary records the outcome plainly: there was no clinically meaningful improvement in dogs on fluoxetine compared with dogs on the control tablet, and because the study did not confirm effectiveness, it was used only for safety data.1
Same drug, different result, and the only thing that changed was the training.
That's why both approved medications are labelled for use in conjunction with a behaviour modification plan rather than on their own.14
Work out which pattern you have. If your dog is anxious most days, you're looking at a daily medication and a behaviour plan. If your dog is fine until something specific happens, you want something fast acting given before the trigger.
Write down the timing for a week before the appointment. When it starts, how long it lasts, what was happening beforehand. That's more useful to me than almost anything else you can bring.
Expect an exam and probably bloodwork first. Pain and thyroid disease both masquerade as anxiety, and these medications are processed by the liver and kidneys, so I want a baseline.
Then give it time. Four to eight weeks on a daily medication, and a proper conversation before you change or stop anything.
You can book online or call us at (519) 896-0532.
There isn't one, and any article that names a single winner is guessing. The right choice depends on whether your dog is anxious most days or only when something specific happens. For daily, ongoing anxiety, fluoxetine and clomipramine have the best evidence and are the only two approved for canine separation anxiety, but both take four to eight weeks. For anxiety you can see coming, like a vet visit, a car trip or a thunderstorm, trazodone or gabapentin work within one to two hours. Plenty of dogs end up on one of each.
It depends entirely on which type you're given. Trazodone, gabapentin and alprazolam work within one to two hours, so you give them before the trigger. Fluoxetine and clomipramine are daily medications that need four to eight weeks before you can judge them. That gap is the single most common reason a medication that would have worked gets abandoned. If you're two weeks into fluoxetine and nothing has changed, that's expected, not failure. Keep a short diary so you're comparing notes rather than memory.
Not without us prescribing it. Some of these are the same molecules we use in dogs, but the doses are not, and human antidepressants and sleep aids are a common cause of poisoning in dogs. Dogs also metabolise several of them differently than people do. On top of that, the human tablets often aren't a size that divides safely for a small dog. If your dog has already swallowed one of your tablets, call us and bring the bottle rather than waiting to see what happens.
It depends on the trigger. For thunderstorms, a placebo controlled trial in 18 storm phobic dogs found gabapentin reduced fear responses. For car travel, a 2026 placebo controlled study found trazodone lowered cortisol, lip licking, restlessness and overall anxiety scores, while gabapentin at 30 mg/kg performed no better than placebo on any measure. So I'd lean towards trazodone for travel and vet visits, and I'd consider gabapentin for storms, especially in an older dog where arthritis pain is part of what's keeping them unsettled.
Many dogs do, and that's a reasonable outcome. Some dogs improve enough with a behaviour plan that we can taper the daily medication after six to twelve months. Others have anxiety severe enough that coming off it means the problem returns, and staying on it is the kinder choice. The one thing I wouldn't do is stop abruptly. Daily medications like fluoxetine and clomipramine get tapered, and we'd normally recheck bloodwork periodically since these are cleared by the liver and kidneys.
Sources
