What Is the Best Medicine for Joint Pain?

For the most common form of joint pain, osteoarthritis, oral anti-inflammatory drugs called NSAIDs consistently outperform every other class of painkiller in head-to-head research. Diclofenac at full prescription doses and etoricoxib have the strongest evidence of clinically meaningful pain relief, and topical diclofenac gels and patches offer a lower-risk alternative for accessible joints like the knee. But “best” depends heavily on what is causing your joint pain, what other health conditions you have, and how long you need treatment, so the real answer is more layered than a single drug name.

NSAIDs Lead the Pack for Osteoarthritis Pain

A large network meta-analysis published in The Lancet compared dozens of NSAID preparations and doses against placebo for knee and hip osteoarthritis. Diclofenac at 150 mg per day and etoricoxib at 60 mg per day had the highest probability of reaching a clinically meaningful level of pain relief, both with a 100% probability of crossing that threshold.¹ A separate BMJ network meta-analysis confirmed that diclofenac 150 mg per day and etoricoxib at both 60 and 90 mg per day had robust evidence of effects stronger than the minimum important difference, while opioids all fell short of that bar.² Pain relief tended to increase with higher NSAID doses, though the dose-response relationship was statistically clear only for naproxen.

Other commonly used NSAIDs like ibuprofen and naproxen also beat placebo, but their effect sizes are smaller at standard over-the-counter doses. That does not mean they are useless. For many people with mild to moderate pain, an over-the-counter dose of ibuprofen or naproxen provides enough relief to get through the day. But if you have tried these and are still struggling, the research suggests that a prescription-strength NSAID, particularly diclofenac or etoricoxib, is more likely to make a noticeable difference.

Acetaminophen Is Weaker Than Most People Assume

Acetaminophen (known as paracetamol in many countries) is still commonly recommended as a first-line option because of its relatively gentle safety profile. But its actual pain-relieving power for osteoarthritis is surprisingly small. A Cochrane review pooling data from randomized trials found that acetaminophen beat placebo by only about 4 points on a 100-point pain scale, an improvement the review authors themselves described as “of questionable clinical significance.”³ It was statistically better than a sugar pill in five of seven trials, but the gap was narrow enough that many patients would not feel a real difference.

This does not mean you should throw away your acetaminophen. If NSAIDs are off limits for you because of kidney problems, stomach ulcers, or heart disease, acetaminophen is one of the few oral options left that carries a low risk of serious harm. It also works well enough for some people, especially in combination with non-drug approaches like exercise or physical therapy. Just don’t expect it to be a powerhouse on its own.

The Safety Trade-Offs With NSAIDs

The reason every doctor does not simply prescribe high-dose diclofenac and call it a day is that NSAIDs come with real risks, and those risks vary by which NSAID you take.

On the cardiovascular side, the largest randomized safety trial ever done for these drugs compared celecoxib, naproxen, and ibuprofen in over 24,000 arthritis patients. Celecoxib was not inferior to either comparator for major cardiovascular events, with event rates of roughly 2.3%, 2.5%, and 2.7% respectively.⁴ A deeper analysis of that same trial found that celecoxib at the doses studied was associated with fewer kidney-related events compared to ibuprofen and showed a trend toward fewer than naproxen.⁵ Naproxen has long carried a reputation as the heart-friendliest NSAID because it is less selective for the COX-2 enzyme, and a pharmacological review supports the idea that this lower selectivity translates to lower cardiovascular risk compared to more COX-2-selective drugs.⁶ That said, the big trial data suggest the gap between naproxen and celecoxib is smaller than once feared.

Stomach and intestinal bleeding is the other major concern. COX-2-selective NSAIDs like celecoxib were originally developed to reduce gut damage, and trial data confirm they cause fewer gastrointestinal ulcers than older NSAIDs in patients not taking aspirin.⁷ However, if you also take aspirin for heart protection, much of that stomach-sparing advantage disappears. A retrospective study found that among aspirin users, the risk of hospitalization for heart attack or GI bleeds climbed substantially with most NSAIDs, with rofecoxib and diclofenac carrying some of the highest adjusted risk.⁸

Kidney damage rounds out the trifecta of NSAID risks. The medical literature links chronic NSAID use to acute kidney injury and, over the long haul, chronic kidney disease. The risk is highest in people who already have cardiovascular or liver problems, or who take multiple medications that affect blood flow to the kidneys.⁹ For short courses in otherwise healthy people, kidney trouble is uncommon. For daily, indefinite use, it is a genuine consideration. And individual genetics play a role too: variations in the CYP2C9 enzyme, which metabolizes several NSAIDs, can slow drug clearance and increase the chance of adverse effects like GI bleeding.¹⁰

Topical Options for Knees and Hands

If you want NSAID-level relief without swallowing a pill, topical diclofenac is the standout option for joints close to the skin surface, particularly knees and hands. The drug absorbs through the skin in small amounts — roughly one-seventeenth the systemic exposure of an equivalent oral dose. A pooled analysis comparing topical and oral diclofenac head to head found that the topical form caused far fewer GI side effects: about 25% of patients on topical reported any GI event versus 39% on oral.¹¹ The trade-off is local skin reactions like dryness and itching, which affected about 29% of topical users.

A meta-analysis comparing different diclofenac formulations found that gels, solutions, and patches all reduced pain and improved function in knee osteoarthritis. Patches provided the strongest short-term pain relief at one to two weeks, while gels and solutions showed sustained benefit over longer periods.¹² The BMJ network meta-analysis also singled out topical diclofenac as the most promising topical treatment, with a high probability of clinically meaningful pain relief.²

Capsaicin cream is another topical option, though it works through an entirely different mechanism. Rather than blocking inflammation, capsaicin overstimulates the nerve endings in your skin until they essentially stop firing pain signals, a process researchers call “defunctionalization” of the nerve fibers. This involves a temporary loss of the nerves’ ability to transmit signals and a reversible retraction of nerve fiber terminals in the skin.¹³ It can take several weeks of regular application before the burning sensation fades and the pain relief kicks in, which many people find hard to stick with.

Injections for When Pills and Creams Are Not Enough

Corticosteroid injections directly into the joint are a mainstay for flare-ups when oral or topical options are insufficient. A meta-analysis found solid evidence for short-term pain improvement lasting up to about two weeks after a knee injection, and the better-designed studies also showed benefits extending out to four to six months.¹⁴ A more recent systematic review showed a similar pattern: a trend toward meaningful reductions in pain and improved function in the first six weeks after injection.¹⁵ However, repeated corticosteroid shots over the long term may cause cartilage damage and accelerate joint degeneration, so most guidelines now advise against frequent or indefinite use.¹⁶

Hyaluronic acid injections, sometimes called viscosupplementation, aim to restore some of the lubricating fluid that deteriorates in arthritic joints. The evidence here is mixed. An umbrella review of 22 systematic reviews found that the vast majority reported statistically significant benefits for pain and function, but when authors applied stricter criteria for clinical relevance, conclusions were split — about three out of five high-quality reviews interpreted the results positively, while the rest were negative or mixed.¹⁷ A systematic review in Cureus concluded that hyaluronic acid injections are best reserved for mild to moderate osteoarthritis in patients who have not gotten enough relief from oral or topical medication. Pain relief tends to last up to about six months regardless of the number of injections or the specific product used.¹⁸

Joint Pain From Gout or Rheumatoid Arthritis Needs Different Drugs

Everything above applies primarily to osteoarthritis — the wear-and-tear kind. If your joint pain is driven by an overactive immune system (rheumatoid arthritis) or by uric acid crystal buildup (gout), the best medicines are quite different.

For gout flares, NSAIDs, corticosteroids, and colchicine are all considered first-line treatments with similar effectiveness; the choice depends on your other health conditions and which side effects you tolerate best.¹⁹ A head-to-head trial comparing naproxen to low-dose colchicine found no meaningful difference in pain relief, but naproxen caused fewer side effects overall. Colchicine triggered diarrhea in nearly half of patients, compared to about one in five on naproxen.²⁰ On the other hand, a study emulating a target trial in over 18,000 gout patients found that NSAID users had a higher rate of major cardiovascular events and cardiovascular death compared to colchicine users when starting urate-lowering therapy.²¹ So for people with gout who also have heart disease risk factors, colchicine may be safer despite the GI discomfort.

For rheumatoid arthritis, the goal shifts from just relieving pain to suppressing the immune attack that is destroying the joint. Disease-modifying drugs like methotrexate are the cornerstone. When those are not enough, biologic therapies targeting specific immune pathways produce substantial improvements. A Cochrane review found that adding a biologic to methotrexate roughly tripled the chance of a major clinical response compared to methotrexate alone.²² Among biologic monotherapies, a network meta-analysis found that tocilizumab achieved greater pain improvements than anti-TNF agents when used alone without methotrexate.²³

Duloxetine for Pain That Does Not Match the X-Ray

Some people with osteoarthritis have pain that seems disproportionate to the visible joint damage. Their nervous system has become sensitized, amplifying pain signals even when inflammation is modest. For these patients, duloxetine — an antidepressant that also modulates pain signaling — is sometimes prescribed. A meta-analysis of five randomized trials found that duloxetine produced a moderate reduction in pain and modest improvement in function over about three months of treatment.²⁴ An individual trial showed significantly greater pain reduction compared to placebo across all time points measured.²⁵

However, a Dutch cluster-randomized trial comparing duloxetine added to usual care versus usual care alone found no difference in knee or hip pain scores at either 3 or 12 months, even among patients specifically screened for centralized pain features.²⁶ The conflicting results likely reflect differences in study design. The positive trials were tightly controlled placebo-blinded experiments; the negative trial was a real-world “pragmatic” study, which tends to produce more modest effects. Duloxetine seems to genuinely help a subset of people, but predicting who will respond remains difficult.

Why Opioids Are a Poor Fit for Joint Pain

Despite their reputation as strong painkillers, opioids perform poorly for osteoarthritis relative to their risks. A systematic review and meta-analysis found only small pain benefits at every time point measured, and stronger opioids actually performed worse than weaker ones while also producing more side effects.²⁷ Increasing doses beyond low levels did not produce meaningfully better pain control but did increase the rate of adverse events. An earlier meta-analysis separately confirmed that even the modest benefits seen in trials came from short studies, and that long-term efficacy and safety data are essentially absent.²⁸ When a systematic review of NSAIDs versus opioids accounted for the people who dropped out of trials due to poor efficacy, the pain reductions were essentially identical across drug classes, meaning opioids offered no advantage while carrying substantially worse side effects and dependence potential.²⁹

Supplements That Have Some Evidence Behind Them

Glucosamine and chondroitin are among the most popular over-the-counter supplements for joint pain, but the evidence is uneven. A meta-analysis of randomized controlled trials found that chondroitin on its own could reduce pain and improve function compared to placebo, while glucosamine’s measurable effect was limited to stiffness improvement. The combination of the two supplements did not clearly beat placebo.³⁰ A separate review noted that glucosamine’s apparent effect has shrunk over time as more and better trials have accumulated, and that the standard daily dose barely reaches therapeutic concentrations in the body.³¹

Turmeric and curcumin extracts have generated more encouraging results in recent years. A systematic review of ten studies found that all showed improvement in pain and function from baseline with turmeric therapy, and three head-to-head comparisons against NSAIDs found no significant difference in outcomes.³² A Bayesian network meta-analysis confirmed that curcumin meaningfully reduced pain scores and overall joint symptoms versus placebo, and that it caused fewer adverse reactions than NSAIDs.³³ These results are promising, but a few caveats apply: curcumin is poorly absorbed without special formulations, the trials tend to be small and short, and the supplement market is not held to pharmaceutical-grade quality standards. Still, for people who cannot tolerate NSAIDs, curcumin extracts are one of the more plausible supplement-grade options.

The Placebo Effect Is Enormous in Joint Pain

One of the most striking findings across joint pain research is how large the placebo response is. Across all studied interventions for knee osteoarthritis, an estimated 75% of the overall treatment effect on pain can be attributed to placebo and contextual factors like the patient’s expectations, the ritual of receiving treatment, and the therapeutic relationship.³⁴ This is not imaginary pain relief — the brain genuinely modulates pain perception, and the expectation of improvement can dampen pain signaling in measurable ways.

The placebo effect varies by how the treatment is delivered. Injections tend to produce larger placebo responses than pills, and more invasive-seeming procedures generate bigger responses than simpler ones.³⁵ This makes it harder to determine the “true” drug effect of injected therapies like hyaluronic acid, and it means that network comparisons between oral and injectable treatments may be unfair if they do not account for the different baseline placebo responses. For you as a patient, this also means that the relationship with your provider, the confidence you have in your treatment plan, and your general expectations all meaningfully shape how much relief you get.

Nerve Growth Factor Blockers on the Horizon

One of the more intriguing drug classes in development targets a protein called nerve growth factor (NGF), which plays a central role in pain signaling from arthritic joints. Tanezumab, an antibody that blocks NGF, produced dramatic pain reductions in early trials, with patients reporting 45 to 62% improvement in knee pain while walking over 16 weeks, compared to 22% with placebo.³⁶ Those are some of the largest effect sizes ever seen in osteoarthritis pain trials. But serious safety problems emerged, including cases of rapidly progressing joint destruction and bone death, which led to a temporary halt of all clinical trials.³⁷ Research has resumed with modified protocols and lower doses, but no anti-NGF drug has yet cleared the regulatory bar for osteoarthritis in most markets. The class represents a genuinely new approach to joint pain, but its future depends on whether the safety risks can be managed without erasing the efficacy advantage.