Evidence-graded guideReviewed October 31, 2026

Herbs and Supplements for Joint Pain and Arthritis: What the Trials Show

The herbs with the best evidence for joint pain are curcumin (matched ibuprofen for knee osteoarthritis in a 367-person trial and shows consistent pain reduction across more than a dozen RCTs), boswellia (significant pain and function gains within 7 days to 12 weeks in several osteoarthritis trials) and omega-3 fish oil, which is the only supplement here with solid data in rheumatoid arthritis, where it reduces tender joints and NSAID use. Ginger, topical capsaicin and devil's claw have moderate support; glucosamine and chondroitin are genuinely contested and we explain why.

This guide is for adults with osteoarthritis (the wear-related kind, usually in knees, hips, hands or spine), with inflammatory arthritis such as rheumatoid arthritis alongside their prescribed treatment, or with non-specific aching joints and stiffness. Those are different diseases: most herbal trials are in knee osteoarthritis, so a pick that is 'Strong' for a sore knee may have no data at all for rheumatoid arthritis. Each entry says which condition the evidence covers.

See a doctor first if a joint is hot, red and swollen (possible infection or gout), if several joints are swollen and stiff for more than an hour in the morning (inflammatory arthritis needs early disease-modifying drugs to prevent damage), if pain follows an injury or comes with fever, weight loss or rash, or if back pain includes numbness, weakness or bladder changes. Supplements can reduce symptoms; none of them stops rheumatoid arthritis from eroding joints.

Quick answer

  • Curcumin — pain relief comparable to ibuprofen in knee osteoarthritis across multiple RCTs
  • Boswellia — reduced pain and improved function in osteoarthritis within 1–12 weeks
  • Omega-3 fish oil — fewer tender joints and less NSAID use in rheumatoid arthritis (2.7 g/day EPA+DHA)
  • Ginger — modest osteoarthritis pain reduction in a meta-analysis of 5 RCTs
  • Capsaicin cream — roughly 50% pain reduction in osteoarthritis after 4 weeks of topical use
  • Glucosamine and chondroitin — mixed: pharmaceutical-grade forms may help, most retail products do not

The 10 best herbs and supplements for joint pain and arthritis

Ranked by the strength of human evidence. Strong = multiple trials or a meta-analysis. Moderate = a few small trials. Preliminary = early or animal data.

  1. Curcumin (turmeric extract)

    Strong evidenceCurcuma longa (isolated curcuminoids)

    Curcumin has the deepest trial base of any herb for osteoarthritis. A 2016 meta-analysis of 8 RCTs found about 1,000 mg/day reduced pain and improved function versus placebo, with effects similar to NSAIDs. In a 367-person trial, 1,500 mg/day of curcumin extract matched 1,200 mg/day of ibuprofen for knee pain and function over 4 weeks with fewer gastrointestinal side effects. A 2021 meta-analysis of 15 RCTs confirmed the pain benefit. Data in rheumatoid arthritis are limited to a few small trials showing reduced disease activity scores.

    Key evidence
    Kuptniratsaikul 2014 RCT vs ibuprofen, 367 patients; Daily 2016 meta-analysis of 8 RCTs
    Studied dose
    500–1,500 mg/day of a bioavailable curcumin extract (with piperine, or a phospholipid or micellar form) for 4–12 weeks
    Best for
    Knee and hip osteoarthritis in people who want to cut NSAID use
    Cautions
    Antiplatelet effect; caution with warfarin, apixaban, clopidogrel and before surgery. Can worsen gallbladder problems and reflux. Piperine-containing products raise levels of many drugs. Rare liver injury reports with high-bioavailability forms; stop if urine darkens.
    Full Curcumin (turmeric extract) profile

  2. Omega-3 fish oil

    Strong evidence

    Fish oil is the only pick with reliable rheumatoid arthritis evidence. A 2017 meta-analysis of 20 RCTs found omega-3 reduced tender joint counts, morning stiffness and NSAID use; the effect needs at least 2.7 g/day EPA+DHA for 12 weeks or more. A 2015 Australian RCT in early RA found high-dose fish oil added to standard therapy doubled remission rates and reduced treatment failure. For osteoarthritis the evidence is thinner: a 2015 trial found low-dose and high-dose fish oil both improved knee pain, with no dose-response.

    Key evidence
    Abdulrazaq 2017 meta-analysis of 20 RCTs in RA; Proudman 2015 RCT in early RA
    Studied dose
    2.7–3 g/day combined EPA+DHA (usually 6–10 standard capsules or a concentrated product) for at least 12 weeks, for rheumatoid arthritis
    Best for
    Rheumatoid and other inflammatory arthritis alongside prescribed DMARDs
    Cautions
    At these doses it adds to anticoagulants and antiplatelet drugs; stop a week before surgery. Can raise LDL slightly and cause loose stools. Choose products tested for oxidation and heavy metals.
    Full Omega-3 fish oil profile

  3. Boswellia (Indian frankincense)

    Moderate evidenceBoswellia serrata

    A 2020 meta-analysis of 7 RCTs (545 patients) found boswellia extract significantly reduced osteoarthritis pain and stiffness and improved function compared with placebo over 4–12 weeks. In one trial, a 5-Loxin extract at 100 mg/day improved pain within 7 days, and a 2019 trial with a 169 mg twice-daily extract showed reduced cartilage breakdown markers alongside pain scores. The evidence is osteoarthritis-specific; rheumatoid data are limited to old, small studies.

    Key evidence
    Yu 2020 meta-analysis of 7 RCTs, 545 patients; Sengupta 2008 RCT of 5-Loxin
    Studied dose
    100–250 mg/day of an extract standardised to 30–65% boswellic acids (AKBA-enriched forms at the lower end), for 8–12 weeks
    Best for
    Knee osteoarthritis with morning stiffness, often combined with curcumin
    Cautions
    May add to anticoagulant effect; stop before surgery. Mild stomach upset, nausea or diarrhoea. May interact with drugs metabolised by CYP enzymes. Not studied in pregnancy.
    Full Boswellia (Indian frankincense) profile

  4. Ginger

    Moderate evidenceZingiber officinale

    A 2015 meta-analysis of 5 RCTs (593 patients) found ginger extract moderately reduced osteoarthritis pain and disability compared with placebo, with a standardised effect of about 0.3 and more withdrawals for mild stomach upset. In the largest trial, 261 people with knee OA took 255 mg of a standardised extract twice daily for 6 weeks: 63% had a pain response versus 50% on placebo. Modest but real; topical ginger compresses have less data.

    Key evidence
    Bartels 2015 meta-analysis of 5 RCTs, 593 patients; Altman 2001 RCT, 261 patients
    Studied dose
    500–1,000 mg/day of a standardised ginger extract (gingerols and shogaols) in divided doses, for 6–12 weeks
    Best for
    Mild knee osteoarthritis, particularly in people who also get nausea or want a gentle option
    Cautions
    Antiplatelet effect; caution with warfarin and before surgery. Heartburn and reflux at higher doses. May lower blood sugar slightly. Safe in pregnancy at food doses but keep supplements under 1 g/day.
    Full Ginger profile

  5. Capsaicin cream (topical)

    Moderate evidenceCapsicum annuum / Capsicum frutescens (source)

    Capsaicin depletes substance P in local pain nerves. A 1991 double-blind trial in 70 osteoarthritis and 31 rheumatoid patients found 0.025% cream four times daily reduced pain by 33% in OA and 57% in RA after 4 weeks versus 21% and 11% with placebo. A 2016 meta-analysis of 5 RCTs in osteoarthritis confirmed a moderate pain reduction. The burning sensation is the main reason people stop, but it fades with use.

    Key evidence
    Deal 1991 RCT, 101 patients; Guedes 2016 meta-analysis of 5 RCTs
    Studied dose
    0.025–0.075% capsaicin cream applied to the painful joint 3–4 times daily for at least 4 weeks
    Best for
    Hand and knee osteoarthritis where you want to avoid oral drugs entirely
    Cautions
    Burning and redness in the first 1–2 weeks is expected; wash hands after use and keep away from eyes and mucous membranes. Do not apply under heat pads or to broken skin. Avoid in people with capsaicin allergy.
    Full Capsaicin cream (topical) profile

  6. Devil's claw

    Moderate evidenceHarpagophytum procumbens

    Devil's claw (Harpagophytum) has a cluster of European trials in osteoarthritis and low back pain. A 2006 Cochrane review of herbal medicines for low back pain found strong evidence that 50–100 mg/day of harpagoside reduced pain more than placebo in the short term, and a 2003 trial found an extract delivering 60 mg/day harpagoside matched the NSAID diacerein in knee and hip OA with fewer side effects. Trials are short (4–16 weeks) and quality is mixed.

    Key evidence
    Gagnier 2006 Cochrane review of herbal medicines for low back pain; Chantre 2000 RCT vs diacerein
    Studied dose
    An extract delivering 50–100 mg/day of harpagoside (typically 600–2,400 mg/day of extract), for 8–12 weeks
    Best for
    Low back pain and osteoarthritis in people wanting a non-NSAID oral option
    Cautions
    Stimulates stomach acid; avoid with ulcers, gallstones and reflux. May affect heart rhythm and blood pressure at high doses. Possible interaction with warfarin and with diabetes drugs. Avoid in pregnancy.
    Full Devil's claw profile

  7. Glucosamine

    Moderate evidence

    Glucosamine is the most contested supplement in medicine. The NIH-funded GAIT trial (1,583 patients) found glucosamine hydrochloride no better than placebo overall, with a possible benefit only in a subgroup with moderate to severe pain. European trials of pharmaceutical-grade glucosamine sulfate 1,500 mg/day reported modest pain relief and slowed joint-space narrowing, while a 2010 network meta-analysis of 10 trials concluded neither glucosamine nor chondroitin had a clinically relevant effect. Industry-funded trials are far more positive than independent ones.

    Key evidence
    Clegg 2006 GAIT trial, 1,583 patients; Wandel 2010 network meta-analysis of 10 trials
    Studied dose
    1,500 mg/day of crystalline glucosamine sulfate (not hydrochloride) as a single dose, for at least 3 months before judging
    Best for
    Knee osteoarthritis in people who want to try a well-tolerated option, accepting that a null result is likely
    Cautions
    Derived from shellfish; check allergy. May slightly raise blood glucose and intraocular pressure. Possible increase in warfarin effect. Not for pregnancy for lack of data.
    Full Glucosamine profile

  8. Chondroitin sulfate

    Moderate evidenceChondroitin sulfate (from bovine, porcine, or avian cartilage)

    Chondroitin fares a little better than glucosamine in recent trials. A 2017 trial of pharmaceutical-grade chondroitin 800 mg/day in 604 knee OA patients found it matched celecoxib and beat placebo for pain and function at 6 months. A 2015 Cochrane review of 43 studies found a small pain benefit (about 8 points on a 100-point scale) of low quality, driven by lower-quality trials. The combination with glucosamine was no better than placebo in GAIT.

    Key evidence
    Reginster 2017 RCT vs celecoxib, 604 patients; Singh 2015 Cochrane review of 43 studies
    Studied dose
    800–1,200 mg/day of pharmaceutical-grade chondroitin sulfate for 3–6 months
    Best for
    Knee osteoarthritis in people who cannot take NSAIDs and are using a quality-tested product
    Cautions
    Mild stomach upset. Possible additive effect with anticoagulants. Retail products vary widely in actual chondroitin content; look for third-party testing. Derived from bovine or shark cartilage.
    Full Chondroitin sulfate profile

  9. MSM (methylsulfonylmethane)

    Preliminary evidence

    A 2017 systematic review found three RCTs of MSM in knee osteoarthritis, all showing small improvements in pain and function at 3–6 g/day over 12 weeks, with the largest (100 patients) reporting a significant WOMAC pain reduction versus placebo. The effect size is small and the trials are short and few. Often combined with glucosamine, where it has not been shown to add anything.

    Key evidence
    Debbi 2011 RCT, 49 patients, 3.4 g twice daily; Brien 2008 systematic review
    Studied dose
    3–6 g/day (often 1.5–3 g twice daily) for 12 weeks
    Best for
    Mild knee osteoarthritis as an add-on to better-studied picks
    Cautions
    Bloating and diarrhoea at higher doses. No known drug interactions, but data are limited. Avoid in pregnancy for lack of data.
    Full MSM (methylsulfonylmethane) profile

  10. Collagen (undenatured type II and hydrolysed)

    Preliminary evidence

    Two different products share the name. Undenatured type II collagen at 40 mg/day beat glucosamine plus chondroitin for knee OA pain and function in a 2016 trial of 191 patients. Hydrolysed collagen peptides at 10 g/day improved activity-related knee pain in athletes in a 24-week Penn State trial. Both are single or few trials, mostly industry-funded, with modest effect sizes.

    Key evidence
    Lugo 2016 RCT of UC-II, 191 patients; Clark 2008 RCT of collagen hydrolysate in athletes
    Studied dose
    40 mg/day of undenatured type II collagen, or 10 g/day of hydrolysed collagen peptides, for 3–6 months
    Best for
    Activity-related knee discomfort in active people; those who have not responded to glucosamine
    Cautions
    Very well tolerated. Check for egg, fish or bovine allergy depending on source. Some products add vitamins; check labels.
    Full Collagen (undenatured type II and hydrolysed) profile

How to choose

  • If you have knee or hip osteoarthritis: start with curcumin, add boswellia if stiffness is prominent, and use capsaicin cream for a single painful joint. Give oral picks 8–12 weeks.
  • If you have rheumatoid or another inflammatory arthritis: fish oil at 2.7 g/day EPA+DHA is the one supplement with solid data, and only alongside your DMARD; nothing herbal replaces methotrexate or biologics.
  • If the main problem is back pain: devil's claw and white willow bark have Cochrane-level evidence for short-term relief; curcumin and boswellia have not been tested specifically for the spine.
  • If you cannot take NSAIDs because of stomach, kidney or heart issues: curcumin, boswellia and topical capsaicin are the sensible substitutes, with ginger as a gentler option.
  • If you want to try glucosamine or chondroitin: use pharmaceutical-grade glucosamine sulfate 1,500 mg or chondroitin 800 mg, commit to 3 months, and stop if nothing has changed. Most cheap combination products contain too little to matter.
  • If a joint is hot, red and swollen, or you have new morning stiffness lasting over an hour, skip this list and see a doctor; gout, infection and early rheumatoid arthritis need specific treatment, fast.

What to skip

Cheap glucosamine-chondroitin combos

The combination was no better than placebo in the 1,583-person GAIT trial, and independent testing has found many retail products contain far less than the label states. If you try these, buy pharmaceutical-grade single agents.

Cat's claw

A single 4-week trial in rheumatoid arthritis and one in knee OA, both small, with no replication in 20 years. It also has immunomodulating effects that could interfere with biologics and transplant drugs.

Bromelain on its own

Trials exist only for combination products with trypsin and rutin, and a 2006 review found insufficient evidence for OA. It is fine as a mild add-on, but not as a primary treatment.

Copper bracelets and magnetic wraps

A 2013 randomised trial of 70 people with rheumatoid arthritis found no effect on pain, inflammation or medication use compared with placebo devices.

Safety and when to see a clinician

  • See a clinician promptly if a joint is hot, red and swollen (infection or gout), if several joints are stiff for more than an hour each morning (possible rheumatoid arthritis, where early DMARD treatment prevents permanent damage), or if joint pain comes with fever, rash, weight loss or new numbness or weakness.
  • Curcumin, boswellia, ginger, fish oil, devil's claw and white willow bark all have antiplatelet or anticoagulant effects. Combine with warfarin, apixaban, clopidogrel or daily aspirin only with medical advice, and stop all of them 1–2 weeks before surgery.
  • Devil's claw, ginger and curcumin can lower blood sugar modestly; diabetes patients on insulin or sulfonylureas should monitor. Devil's claw also stimulates stomach acid and is unsuitable with ulcers or gallstones.
  • Curcumin products with piperine raise blood levels of many drugs, and high-bioavailability curcumin has rare liver injury reports. Stop with dark urine, pale stools or yellowing eyes.
  • Glucosamine is shellfish-derived and may raise blood glucose and eye pressure; chondroitin and MSM have limited safety data in pregnancy. None of the oral picks is recommended in pregnancy without advice.
  • Keep taking prescribed disease-modifying drugs. Supplements can ease symptoms of rheumatoid arthritis but none stops joint erosion, and stopping methotrexate or a biologic to go natural is the single most damaging decision people make.

What matters as much as any herb

  • Weight loss: in knee osteoarthritis each kilogram lost removes about 4 kg of load per step, and a 10% loss of body weight reduced pain by roughly half in the IDEA trial. It outperforms every supplement on this page.
  • Strength training: quadriceps and hip strengthening reduce knee OA pain by an amount comparable to NSAIDs in meta-analyses, and the benefit lasts as long as you keep training.
  • Movement rather than rest: for osteoarthritis and back pain, daily walking, cycling or pool exercise reduces stiffness; prolonged rest worsens both. Aim for 150 minutes a week of something you can sustain.
  • Topical NSAIDs: diclofenac gel gives most of the pain relief of oral NSAIDs for hand and knee OA with a fraction of the stomach and kidney risk, and combines well with the oral herbs here.
  • Sleep and mood: poor sleep amplifies pain sensitivity, and depression roughly doubles reported arthritis pain. Treating either changes pain scores more than most supplements.

Frequently asked questions

What is the best natural anti-inflammatory for joints?

Curcumin has the strongest trial record: in a 367-person trial, 1,500 mg/day matched ibuprofen for knee osteoarthritis pain with fewer stomach side effects, and meta-analyses of 8–15 RCTs confirm the benefit. Boswellia is the usual partner, with faster onset in some trials.

Does turmeric really help arthritis?

Turmeric powder from the spice rack contains only about 3% curcumin and is poorly absorbed, so the trials used concentrated, bioavailable extracts at 500–1,500 mg/day. At those doses, yes, osteoarthritis pain improves by an amount similar to NSAIDs over 4–12 weeks. Rheumatoid data are much thinner.

Does glucosamine work for joint pain?

It depends on the form and who you ask. The large NIH GAIT trial found glucosamine hydrochloride no better than placebo, while European trials of pharmaceutical-grade glucosamine sulfate 1,500 mg/day show modest benefit. Independent meta-analyses conclude the effect, if any, is small. Try sulfate for 3 months, then decide.

What helps rheumatoid arthritis naturally?

Fish oil at 2.7 g/day or more of EPA+DHA reduces tender joints, morning stiffness and NSAID need across 20 RCTs, and doubled remission rates when added to standard therapy in early RA. Curcumin has a few small positive trials. All of this is alongside DMARDs, never instead of them.

How long do joint supplements take to work?

Boswellia showed benefit within 7 days in one trial and most curcumin trials ran 4–8 weeks. Fish oil in rheumatoid arthritis needs at least 12 weeks, and glucosamine, chondroitin and collagen need 3 months. Capsaicin cream takes 2–4 weeks as the initial burning fades.

Is it safe to take curcumin with blood thinners?

Curcumin has antiplatelet activity and there are case reports of raised INR with warfarin, so combining them needs medical supervision and monitoring. The same applies to boswellia, ginger, fish oil and devil's claw. Stop all of them 1–2 weeks before surgery.

Which is better for joints, collagen or glucosamine?

In the one head-to-head trial, 40 mg/day of undenatured type II collagen beat glucosamine plus chondroitin for knee osteoarthritis pain and function over 6 months. It is a single industry-funded trial, so the honest answer is that both are modest options and curcumin outranks either.

References

  1. Kuptniratsaikul V, et al.. Efficacy and safety of Curcuma domestica extracts compared with ibuprofen in patients with knee osteoarthritis: a multicenter study. Clinical Interventions in Aging (2014)View source
  2. Daily JW, Yang M, Park S. Efficacy of turmeric extracts and curcumin for alleviating the symptoms of joint arthritis: a systematic review and meta-analysis of randomized clinical trials. Journal of Medicinal Food (2016)View source
  3. Yu G, et al.. Effectiveness of Boswellia and Boswellia extract for osteoarthritis patients: a systematic review and meta-analysis. BMC Complementary Medicine and Therapies (2020)View source
  4. Abdulrazaq M, Innes JK, Calder PC. Effect of omega-3 polyunsaturated fatty acids on arthritic pain: a systematic review. Nutrition (2017)View source
  5. Proudman SM, et al.. Fish oil in recent onset rheumatoid arthritis: a randomised, double-blind controlled trial within algorithm-based drug use. Annals of the Rheumatic Diseases (2015)View source
  6. Bartels EM, et al.. Efficacy and safety of ginger in osteoarthritis patients: a meta-analysis of randomized placebo-controlled trials. Osteoarthritis and Cartilage (2015)View source
  7. Clegg DO, et al.. Glucosamine, chondroitin sulfate, and the two in combination for painful knee osteoarthritis. New England Journal of Medicine (2006)View source
  8. Wandel S, et al.. Effects of glucosamine, chondroitin, or placebo in patients with osteoarthritis of hip or knee: network meta-analysis. BMJ (2010)View source
  9. Singh JA, et al.. Chondroitin for osteoarthritis. Cochrane Database of Systematic Reviews (2015)View source
  10. Gagnier JJ, et al.. Herbal medicine for low back pain. Cochrane Database of Systematic Reviews (2006)View source
  11. Deal CL, et al.. Treatment of arthritis with topical capsaicin: a double-blind trial. Clinical Therapeutics (1991)View source
  12. Turmeric: Usefulness and Safety. National Center for Complementary and Integrative HealthView source

Related conditions

More guides

Not sure which one fits you?

Ask AskMN. It knows these studies, checks interactions with the medications you tell it about, and it is free and unlimited, no account needed.

Ask about joint pain and arthritis

This guide is educational and is not medical advice. Herbs and supplements can interact with prescription medicines and are not regulated like drugs. Talk with a clinician or pharmacist before starting anything new, especially if you are pregnant, breastfeeding, have a chronic condition, or take medication.