Dog Arthritis Medication: What Your Vet Is Actually Choosing Between

The diagnosis is on the discharge note, and somewhere under it is a line about starting an anti-inflammatory. So you searched for what that means, and the results came back as rankings: ten best arthritis medications, eight best with top picks for 2026, a pharmacy comparison, a retailer's category page.

Read past the headlines and those same pages concede the point. No dog arthritis medication is best in the abstract, because the choice is not made by comparing the drugs to each other. It is made by comparing each drug against your particular dog — her kidneys, her liver, her stomach, her age, everything else she takes. The drug list is short and public; the constraints are what the appointment is about.

Below: what is genuinely on the menu, what the one published head-to-head trial found and why it settles less than it appears to, what long-term safety monitoring actually involves, the open question hanging over the newest option, and how to tell whether any of it is working.

What medications are actually used for arthritis in dogs?

Four groups, and only the first is a first-line treatment. Merck's Veterinary Manual is direct about the hierarchy: "NSAIDs (including the subclass of prostaglandin E2 type 4 [EP4] receptor antagonists) are the most predictably effective treatment for OA."³ Everything else is added around them, or used when they cannot be.

VCA names the same actives a vet chooses among: "meloxicam (Metacam® and other trade names), carprofen (Rimadyl®), deracoxib (Deramaxx®), grapiprant (Galliprant®), robenacoxib (Onsior®), and firocoxib (Firodyl®)."⁴ The FDA's approved list matches it.¹ Every one of them is prescription-only — the FDA states that "all approved veterinary NSAIDs are only available by prescription"⁹ — which is worth sitting with: the retailer and pharmacy pages ranking for this search describe products you cannot obtain by deciding you want them. The same FDA page notes that veterinarians "should give these sheets to owners with every NSAID prescription," meaning the drug-specific Client Information Sheet is something you can ask for by name if it does not arrive in the bag.⁹

Notice the shape of that last column: nothing in it is a score. Each row passes or fails on something about the individual dog — which is why a list ranked by anything else cannot tell you where your dog belongs in it.

Which arthritis medication is best for a dog?

Nobody has published an answer, and the closest thing to one comes with two disclosures attached. The comparative evidence the ranking posts imply mostly does not exist, and the clearest statement of that comes from the one team who went looking: "to date there have been no efficacy studies comparing grapiprant with other NSAIDs," they wrote, noting that "the only report of grapiprant efficacy has been of a field study in which it was demonstrated to be superior to placebo."⁵ Superior to nothing is not the same as superior to the alternative, and it is the evidence most of these drugs carry.

One study did ask it. García de Salazar Alcalá and colleagues compared firocoxib and grapiprant in 18 healthy Beagles, inducing acute arthritis with an injection of sodium urate crystals and measuring how much weight each dog would put back on the limb. The result was lopsided. Across the first experiment firocoxib held lameness ratios at or above 0.80 at every timepoint, while grapiprant sat at zero at five and seven hours. The authors wrote that "the reduction in lameness provided by firocoxib was consistently superior to that provided by grapiprant, which was not significantly different from untreated controls."⁵

Now the disclosures, both of which are in the paper itself. It states that "the study was funded by Merial (now part of Boehringer Ingelheim), the manufacturer of Previcox®" — the firocoxib product that won. And the authors qualify their own finding: "based on the mode of action, which could be more progressive for grapiprant than for firocoxib, this model of acute pain control could not be sufficient to show pain and anti-inflammatory control under chronic conditions like osteoarthrosis."⁵

That is not a reason to dismiss the result, but a reason to see what it is: one manufacturer-funded single-dose experiment, in an artificial acute injury, in young laboratory Beagles, whose authors say it may not measure the thing grapiprant is prescribed for. Grapiprant is an approved drug for dogs in its own right.¹

One marketing claim does not survive the same paper, though. Grapiprant is widely promoted — including on pages ranking for this search — as the NSAID with meaningfully less risk attached. The authors state flatly that "there is no clinical evidence of any safety advantages over other NSAIDs."⁵ A different mechanism is a plausible reason to expect a safety advantage. It is not a demonstrated one.

So the honest summary is uncomfortable for a ranked list and useful in an exam room: the drugs have not been meaningfully raced against one another in the disease they treat. What your vet is choosing between is not first place and second. It is which of six roughly comparable options your dog's organs, history and other prescriptions leave available.

How safe are NSAIDs for a dog who takes them for years?

Safe enough to be the default, and monitored precisely because they are not risk-free. The FDA's guidance to veterinarians is specific about what has to happen before the first tablet: "All dogs should undergo a thorough history and physical examination before initiation of NSAID therapy. Appropriate laboratory tests to establish baseline blood values prior to, and periodically during, the use of any NSAID are strongly recommended."¹

The reason is organ-level. The serious end of the range runs to bleeding ulcers, perforations of the stomach or intestine, and kidney and liver problems.¹ Because osteoarthritis is, in Merck's words, "a lifetime disease," that baseline check is not a one-off.³ The FDA recommends "regular check-ups and periodic blood and urine tests … to determine if the drug is safe for continued use."²

Between those appointments, the monitoring is yours. The FDA gives owners a mnemonic for the signs that should stop the drug, and it is worth taping inside a cupboard door.

The FDA's owner sheet lists two further signs alongside those four: yellowing of the gums, skin or whites of the eyes, and a change in drinking.¹ Both are worth the same phone call. None of this argues against an NSAID — it is why the class can be used for years at a time.

What should you know about Librela before the first injection?

That it is a genuine option with a genuine open question, and that your vet should hear you ask about both. Bedinvetmab was approved by the FDA on 5 May 2023 "for the control of pain associated with osteoarthritis (OA) in dogs," given "by subcutaneous injection once a month and dosed by weight range to target a minimum dose of 0.5 mg/kg."⁷ It is an anti-nerve growth factor monoclonal antibody, which means it works nothing like an NSAID — a real advantage for a dog whose kidneys or stomach rule the tablets out. Merck recommends it "for dogs with moderate to severe clinical signs from OA (COAST stages 3 and 4)."³

Then there is the surveillance. On 16 December 2024 the FDA issued a Dear Veterinarian Letter reporting that its Center for Veterinary Medicine had completed an evaluation of adverse events reported in treated dogs, including "ataxia, seizures, other neurologic signs, including but not limited to, paresis, recumbency, urinary incontinence; polyuria, and polydipsia," and that "in some cases, death (including euthanasia) was reported as an outcome of these adverse events." The letter also notes something about the FDA's own powers that surprises most owners: it "does not currently have the authority to mandate safety-related labeling changes."⁶ Separately, a 2025 commentary in Frontiers in Veterinary Science describes a disproportionality analysis of the European Medicines Agency's EudraVigilance database reporting "a significantly elevated rate of serious musculoskeletal adverse events (MSAEs) — including ligament and tendon injuries, polyarthritis, fractures, musculoskeletal neoplasia, and septic arthritis — in dogs treated with bedinvetmab compared to six other osteoarthritis medications."⁸

What that does and does not mean matters. Spontaneous-reporting databases collect what people choose to report; a disproportionality analysis finds patterns in those reports. Neither can establish that a drug caused an event, and older arthritic dogs are, separately, dogs who fall, tear ligaments and develop tumours. The signal is real and unresolved; causation is not demonstrated. The reasonable position is neither "avoid it" nor "the regulators would have pulled it" — it is to ask your vet directly what she has seen, why she is proposing it for your dog specifically, and what would make her stop.

Can a dog take two arthritis medications at once?

Two NSAIDs, never. Drugs from different classes, often — and that distinction is the single most useful thing to take from this page. The FDA states the rule without qualification: "Only one NSAID should be administered to a dog at any given time," and NSAIDs "should not be combined with a corticosteroid."² Its owner guidance is blunter still: "Never give aspirin or corticosteroids (such as prednisone) along with an NSAID to your dog."¹ VCA puts the same boundary around steroids: the two classes "CANNOT be used together."⁴

Switching between NSAIDs is subject to the same logic. The FDA defines what usually sits in between: "a wash-out period is a period of time during which the dog does not receive any NSAID."² How long it runs is a clinical judgement, which is why "the last one stopped working, we tried something else from the cupboard" is a genuinely dangerous sentence.

Combining across classes is different, and it is how most long-running plans end up looking. VCA describes gabapentin as a drug that "addresses chronic pain differently from NSAIDs or corticosteroids and complements those medications," while noting that "other less commonly used drugs include tramadol, amantadine, and acetaminophen, but they are less effective unless paired with other drugs."⁴ Merck is similarly measured about amantadine: "one study demonstrates the possible utility of amantadine (3–5 mg/kg, PO, every 12–24 hours) as an adjunctive treatment."³ Those figures are what the guidelines record, not a dose for you to act on.

The practical risk in a multi-drug plan is not exotic. It is a daily tablet given twice because two people each thought the other had not — which for an NSAID is exactly the class where doubling matters. If that has just happened with meloxicam, our guide to what to do in the next hour after two doses of Metacam covers it.

Do joint supplements and Adequan actually do anything?

One of those two has a systematic review against it, and the other has a dose. They are shelved together and routinely discussed as one category, which does both a disservice.

On glucosamine and chondroitin, Merck reports that a systematic review "found no evidence of a beneficial effect and concluded that the available evidence did not support the use of these supplements for pain management in OA in dogs and cats."³ That is about as clear as veterinary evidence summaries get, and it covers the most heavily marketed products in the aisle.

Polysulfated glycosaminoglycan is a different kind of product — a prescription injectable, not a chew. Merck records it for dogs "(PSGAG; 4.4 mg/kg, IM or SC, twice weekly for up to 8 treatments)" and notes it "may be more beneficial in the earlier stages of OA."³ Omega-3 fatty acids sit somewhere between: VCA lists them among the common joint products without the same evidence caveat Merck attaches to glucosamine.⁴ We went through the dog joint-supplement claims one at a time in our review of dog supplement trends.

How do you tell whether the medication is working?

By what your dog does on ordinary days, recorded before you forget — because the thing you are measuring changes too slowly to remember accurately. Osteoarthritis pain moves over weeks, and the drug is usually started at the same appointment as three other changes, which makes "is it helping?" genuinely hard to answer six weeks later from memory.

It is worth knowing which lever has the most evidence behind it, and it is not on the prescription. Merck states that "the primacy of even very modest weight loss to improving function in overweight dogs with OA is well established."³ If your dog is carrying extra weight, that is the intervention with the firmest footing — our guide to working out whether your dog is overweight covers body condition scoring. Treatment is a package, not a tablet.

What to bring back to the recheck, in the order a vet will ask for it:

1. Every dose actually given — including the missed ones, which are data rather than confessions. A drug judged ineffective at 80% adherence is a different problem from one judged ineffective at 100%.
2. Stairs, jumps and the sofa — the days she managed them and the days she did not, rather than an overall impression.
3. Where the walk turns — how far in she slows, and whether that distance is moving.
4. Nights — whether she settles or shifts and repositions.
5. Weight, on the same scales — the one number with the firmest evidence behind it.
6. Anything from the BEST FRIEND list, with the date it started.

Our guide to spotting and tracking arthritis in dogs and cats sets out which signs are worth logging and how to read the trend. Take that to the appointment and the conversation changes from an impression to a record, which is what a decision about continuing, changing or adding a drug actually needs.

Veterinary medical disclaimer: SteadyTails is a logging and coordination tool, not a substitute for veterinary care. Nothing on this page is a diagnosis, a prescription or a dosing instruction. Every medication named here is prescription-only, and the doses quoted are figures recorded in veterinary references, not amounts to give your dog. Which drug suits your dog, at what dose, alongside what else, and for how long are decisions that depend on her examination and bloodwork and can only be made by your veterinarian. Never give your dog a human pain reliever, a medication prescribed for another animal, or a second NSAID, and never combine an NSAID with aspirin or a steroid. If your dog shows behaviour changes, stops eating, develops skin redness or scabs, or has black or bloody stool, diarrhoea or vomiting while taking an NSAID, stop the drug and contact your veterinarian. If she collapses, will not bear weight at all, or has been given a double dose, contact your veterinarian or an emergency animal hospital straight away; in the US, ASPCA Animal Poison Control is available around the clock on 888-426-4435.