Ashwagandha and Exercise: What the Research Actually Shows

The name is usually translated as the smell of a horse. Tradition has long read something more ambitious into it: the strength of one. That reading has now travelled a long way from its origins, out of the Ayurvedic texts and into the gym bag, where ashwagandha sits beside the protein tub and the creatine.

The gym is a useful place to test a claim. Strength and stamina can be measured to the kilogram and the second. So the question is simple enough. When researchers put this root in front of people who train, and then measure what those people can do, what happens?

This year, two independent teams tried to answer it properly.

What an Exercise Study Actually Measures

Performance is not one thing, and a supplement can touch one part of it while leaving the rest alone.

Researchers usually separate four capacities. Aerobic endurance, measured as VO2max, the most oxygen the body can use, or as time to exhaustion on a treadmill or bike. Maximal strength, usually a one-repetition maximum, the heaviest single lift. Power, the explosive kind, measured with jumps and throws. And muscular endurance, how many repetitions you can manage before failure.

When studies are pooled, the results are expressed as an effect size. By the usual rule of thumb, 0.2 is small, 0.5 is moderate and 0.8 is large. Keep those three numbers in mind. They are the whole grammar of what follows.

The Big Picture: Thirteen Trials

The broader of the two reviews was published in Nutrients in June 2026 by Li and colleagues. They searched five databases up to April 2026 and found 13 randomised controlled trials comparing ashwagandha with placebo or a control, covering 599 participants and 79 separate performance measurements. Most participants were healthy young adults or athletes.

Averaged across everything, ashwagandha was associated with better performance than placebo, an effect size of 0.47. Small to moderate. The authors rated their confidence in that figure as moderate.

Then comes the number that matters more. A meta-analysis can also estimate a prediction interval, the range of results a new trial of the same kind would be expected to find. Here it ran from minus 0.40 to 1.33. It crosses zero. In plain terms, the next comparable study could find a clear benefit, or nothing at all, or a result slightly in the wrong direction.

An average is not a promise. The honest reading of this literature lives in the gap between the two.

Endurance, Not Explosiveness

The more useful finding sits inside the averages. The type of exercise was the clearest single factor in whether ashwagandha appeared to do anything.

Aerobic endurance showed the most stable result, an effect size of 0.54, with a confidence interval that stayed clear of zero and a moderate certainty rating. Strength had a positive estimate of 0.48, but the interval dipped below zero, so it did not reach statistical significance in this analysis. Power showed no clear improvement. Muscular endurance was too imprecise to call.

Trained people showed a larger estimate than untrained people, 0.65 against 0.33, though the difference between the two groups was not statistically significant. Most of the trials used 500 or 600mg a day of extract, and that is the range the authors judged best supported. The apparently larger effect at 1000mg rested on too few data points to mean much.

The authors were blunt about the foundations. Of the 13 trials, one was rated at low risk of bias. Nine raised some concerns and three were rated high risk. The median statistical power of the studies was estimated at 14 per cent, which is a technical way of saying most of them were too small to detect a modest effect reliably. Their own verdict was that the pooled effect should be read as provisional rather than definitive.

The Strength Question

Strength is where the marketing is loudest, so it deserves a closer look.

The second review, published in Nutrients in August 2026 by Lee and Heo, asked a narrower question. Does ashwagandha add anything to a structured resistance-training programme? They searched twelve bibliographic sources and two trial registers, and found only four randomised trials that fitted. Three could be pooled, covering 161 healthy adults on 500 to 600mg a day of a standardised extract for eight to twelve weeks.

On first reading, the numbers look good. Ashwagandha was associated with larger gains than placebo in bench-press one-repetition maximum, an effect size of 0.60, and in lower-body one-repetition maximum, 0.52. Chest circumference rose more too, at 0.86, though that came from only two trials. On body fat, there was no significant difference.

On second reading, the picture thins. When the authors applied a statistical correction designed for small numbers of studies, both strength results lost their significance. Each depended on a single influential trial. The certainty of evidence was rated low for strength, and lower still for chest circumference. All three pooled trials were funded or supplied by ashwagandha manufacturers. The review itself received no external funding.

Their conclusion does not soften any of this. Current evidence, they wrote, does not establish that adding ashwagandha meaningfully enhances the adaptations produced by resistance training. There is a possible benefit for maximal strength. It is small in evidence and fragile in statistics.

One detail from the paper stays with you. A fourth trial met every inclusion criterion but could not be pooled. It was a preprint, not yet peer reviewed, written by employees of the manufacturer, and its chart placed the average starting bench press at roughly 5.9kg. As the reviewers observed, that is less than the weight of an empty barbell. Careful reading is not pedantry. Sometimes it is the only thing standing between a claim and a customer.

A third 2026 review, by Zhu and colleagues in Frontiers in Pharmacology, pooled 20 trials across cognitive and physical outcomes. It found strength gains concentrated in physically active people, and no significant gain in untrained ones. Consistent, as far as it goes, with the idea that whatever ashwagandha does, it does alongside training rather than instead of it.

A Trial That Complicates the Dose

One more study cuts across the dose story.

In March 2026, Prajapati and colleagues published an eight-week, double-blind, placebo-controlled trial in Phytotherapy Research using just 30mg a day of a branded, root-only extract, concentrated to more than 15 per cent withanolides. Against placebo, the extract was associated with improved VO2max, lower lactic acid during exercise and lower ratings of perceived exertion and fatigue. No treatment-related adverse events were reported.

Two cautions. VO2max was estimated indirectly from a treadmill protocol rather than measured directly from breath gases, and this is one trial of one proprietary extract, not yet replicated. The point it makes is a different one. Thirty milligrams of one preparation and six hundred of another are not the same thing expressed at different volumes. The milligram figure on its own says almost nothing.

Why It Might Work, and Why It Is Not a Pre-Workout

Ashwagandha contains no caffeine and is not a stimulant. Nobody in this literature proposes that it works the way a pre-workout does, on the day, within the hour.

The explanation the researchers favour is slower and less direct. Li and colleagues describe it as a chronic recovery and adaptation aid rather than an acute performance aid. The suggested pathways run through the stress response, sleep, inflammation and oxygen use, the same territory covered in our pieces on ashwagandha and cortisol and ashwagandha and sleep. If there is an effect on the track or the bar, it probably arrives by way of the hours between sessions.

That also explains the shape of the results. Endurance is built across weeks of consistent training and recovery. A single explosive jump is not. It would be odd if a slow-acting root showed up most clearly in the fastest movements.

For anyone who competes under anti-doping rules, the same review makes a practical point. Any supplement carries a contamination risk, and third-party tested products reduce it.

The Caveat That Applies to All of It

The trials did not all test the same thing. They used different branded extracts, different plant parts, different extraction methods and different withanolide contents. Li and colleagues could not identify any single form or brand as better than another, because formulation was tangled up with dose, duration and training programme. A result obtained with one extract at one dose does not automatically transfer to another. It is the recurring lesson of this ingredient, and the one most often missing from the label.

Safety, and the UK Review Underway

An honest write-up of this ingredient does not stop at the benefits.

In trials, ashwagandha is generally well tolerated over a few months. One large recent safety study, published in Phytotherapy Research in 2026, randomised 1,002 adults with stress and anxiety to 600mg a day of root extract or placebo for eight weeks. It reported no serious adverse events and no significant changes in liver function, kidney markers or blood counts compared with placebo. Evidence over longer periods is still lacking, and there are documented case reports of more serious harm, including liver injury. It may also interact with medication for the thyroid, for diabetes and for blood pressure, and with sedatives and immunosuppressants.

Regulators are paying attention. Denmark has restricted ashwagandha in food supplements. France, through its food safety agency ANSES, has advised against use in certain groups, including pregnant and breastfeeding women. In 2026 Health Canada completed a safety review of ashwagandha and liver injury, concluding that it did not find sufficient evidence to establish a definite link, but that a possible link could not be ruled out. It will continue to monitor.

In the United Kingdom, the Food Standards Agency ran a call for evidence on ashwagandha between 8 July and 1 September 2024, and published a summary of the responses on 25 June 2026. Those responses go to the Committee on Toxicity, and a risk assessment will follow, with the outcome to be published. As things stand, no safe consumption level has been established in the UK and no final conclusion has been reached. Ashwagandha is not banned here and remains legally available, but anyone taking it should know that a review is genuinely in progress.

If you are pregnant or breastfeeding, under eighteen, have a thyroid or liver condition, or take any prescribed medication, this is a conversation to have with your doctor or pharmacist rather than a decision to make from a blog post.

The Honest Summary

Across a growing number of trials, ashwagandha is associated with modestly better exercise performance, most consistently in aerobic endurance, at around 500 to 600mg a day of a standardised extract, taken over weeks rather than hours. The strength story is plausible and the direction is consistent. It is also resting on three small, industry-funded trials that do not survive the strictest statistical test.

The horse in the name was always a metaphor. The research so far suggests something closer to steady stamina than sudden strength, and it suggests even that with care.

The training still does the work.

References

1. Li, X., Li, H., Yao, S., Hou, Y. and Chi, A. Effects of Withania somnifera (ashwagandha) supplementation on exercise performance, a systematic review and three-level meta-analysis. Nutrients, 18(12), 1915, 2026. DOI
2. Lee, J.C. and Heo, A.S. Adjunctive ashwagandha (Withania somnifera) supplementation and resistance-training adaptations in healthy adults, a systematic review and random-effects meta-analysis of muscular strength, chest circumference, and body composition. Nutrients, 18(17), 2815, 2026. DOI
3. Zhu, X., Zeng, Q., Lei, Y. and Xu, D. Effects of Withania somnifera (ashwagandha) on cognitive and physical function in adults, a systematic review and meta-analysis. Frontiers in Pharmacology, 17, 1799467, 2026. DOI
4. Prajapati, H. et al. Efficacy and safety of low-dose ashwagandha supplementation on exercise endurance, a randomized, placebo-controlled, double-blind, clinical trial. Phytotherapy Research, 40(8), 4978 to 4987, 2026. DOI
5. Pakhale, K. et al. Safety of 8-week administration with ashwagandha (Withania somnifera) root extract in adults with stress and anxiety, findings from a prospective, randomized, multi-center, double-blinded, placebo-controlled study. Phytotherapy Research, 40(8), 5005 to 5015, 2026. DOI
6. Health Canada. Health Product InfoWatch, August 2026: summary safety review of ashwagandha and the potential risk of liver injury. Link
7. National Institutes of Health, Office of Dietary Supplements. Ashwagandha, health professional fact sheet. Link
8. Food Standards Agency. Summary of responses to call for evidence on ashwagandha, published 25 June 2026. Link

This article is for informational purposes only and is not intended as medical advice, nor as a description of any particular product. It discusses the botanical Withania somnifera and published third-party research. Food supplements are not a substitute for a varied, balanced diet and a healthy lifestyle. Individual responses vary. Ashwagandha is not suitable for everyone, including during pregnancy and breastfeeding, and it may interact with prescribed medication. Always consult a qualified healthcare professional before starting or changing any supplement routine, particularly if you are taking medication or have a diagnosed health condition.