Ashwagandha (Withania somnifera) is the most-studied adaptogen in the supplement aisle, and it is one of the rare herbal products where honest coverage is genuinely favorable. Unlike many botanicals, it has a real stack of randomized, placebo-controlled human trials — and its strongest, most reproducible signal is exactly the one it’s marketed for: lowering stress and the stress hormone cortisol. The catch is that the trials are mostly small, short, and often funded by the companies that sell the standardized extracts, and there is a real — if rare — liver safety signal that deserves a clear-eyed mention. Here is what the evidence actually supports, tier by tier.
Stress and cortisol: the strongest signal
This is where ashwagandha earns its reputation. In a frequently-cited randomized, double-blind, placebo-controlled trial, adults with a history of chronic stress took a high-concentration root extract (300 mg twice daily) for 60 days and saw substantially larger reductions in perceived stress scores and in serum cortisol than placebo, with cortisol falling roughly a quarter to a third from baseline.[1] A separate trial of a different standardized extract (Shoden, 240 mg/day) replicated the core finding: reduced anxiety and a significant drop in morning cortisol versus placebo over 60 days.[2] A third RCT in healthy but stressed adults again showed dose-related reductions in perceived stress and cortisol.[3] Three independent trials, two different commercial extracts, the same direction of effect — that consistency is what separates ashwagandha’s stress claim from typical supplement hype.
The honest caveats: these are small studies (typically 50–120 people), they run for weeks rather than months, they lean on self-reported stress scales, and most were funded by or conducted with the extract manufacturers. None of that makes the results fake — the effect sizes are real and repeated — but it means the appropriate framing is “a well-supported short-term effect on stress and cortisol,” not “a proven long-term treatment.”
Anxiety and sleep: decent, moderate evidence
The anxiety data travel alongside the stress data — the same trials that measured cortisol also measured anxiety scales and found improvements.[2][3] Sleep is a step less certain but still supported: a randomized, placebo-controlled study in people with insomnia and anxiety found that root extract (300 mg twice daily) for eight weeks improved sleep onset, sleep quality, and sleep efficiency versus placebo.[4] The effect is meaningful but moderate, and the trials are smaller than you’d want. If you’re reaching for ashwagandha, stress relief and sleep quality are the two things the evidence best supports it doing.
Testosterone and strength: a modest, thinner signal
This is the claim that gets oversold. There is a signal — it is just small and built on few trials. In resistance-trained men, ashwagandha (KSM-66, 300 mg twice daily) alongside a strength-training program produced greater gains in muscle strength and size, better recovery, and a rise in testosterone compared with placebo.[5] A separate crossover trial in aging, overweight men reported a modest increase in testosterone and DHEA-S over eight weeks.[6]Both are small, and the testosterone changes are modest — a nudge, not the several-hundred-point swings you get from prescription therapy. Ashwagandha is not a substitute for treating clinical hypogonadism; for that, see our review of testosterone replacement therapy. If you’re optimizing naturally, it belongs in the same “modest lever” tier as tongkat ali and the everyday inputs covered in foods that support testosterone.
Which extract? KSM-66 vs Shoden
Standardization is not marketing fluff here — it’s the reason the trials are interpretable. Almost all the good data are on two branded root extracts. KSM-66 is a root-only extract standardized to about 5% withanolides, studied mostly at 300–600 mg/day, and is the extract behind much of the stress, sleep, and strength data.[1][3][5] Shoden is a more concentrated extract standardized to roughly 35% withanolide glycosides, dosed lower (about 240 mg/day), used in the cortisol and anxiety trials.[2] The practical point: a generic “ashwagandha root powder” off a shelf has not been shown to do what these standardized extracts did. If you want the studied effect, use the studied ingredient at the studied dose.
Safety: well-tolerated, but a real liver signal
In the trials above, ashwagandha was generally well tolerated, with side effects (mild GI upset, drowsiness) close to placebo over the weeks studied. But the honest safety picture includes a genuine red flag: a published case series from Iceland and the US Drug-Induced Liver Injury Network documented ashwagandha-induced liver injury — cholestatic or mixed hepatotoxicity with jaundice, nausea, and itching, typically appearing 2–12 weeks after starting, and resolving after people stopped.[7] These cases are rare relative to how many people use the supplement, and most fully recovered — but they are real, biologically plausible, and a reason to stop immediately and see a clinician if you develop dark urine, jaundice, or right-upper-abdominal discomfort. People with existing liver disease, and anyone pregnant (traditionally avoided) or on thyroid or sedative medications, should treat ashwagandha as a drug worth clearing with a clinician, not a harmless herb.
Timing and how to use it sensibly
There’s no strict clock on ashwagandha, but the trials offer sensible defaults: it was usually taken with food, split into a morning and evening dose (for KSM-66, 300 mg twice daily), and given time — benefits accrued over 4–8 weeks, not on day one. Because it can be mildly sedating and the sleep data are favorable, many people prefer the larger dose in the evening. If stress or sleep is the goal, an 8-week trial at a studied extract and dose is a reasonable, honest test; if you feel nothing after that, it likely isn’t your lever.
The honest bottom line
Ashwagandha is a rare supplement where the marketing and the evidence mostly point the same way — for its main claim. The human data are strongest for reducing stress and cortisol, decent for anxiety and sleep, and modest and thinner for testosterone and strength. Use a standardized extract (KSM-66 or Shoden) at the studied dose, understand that most trials are small, short, and industry-linked, and respect the rare-but-real liver-injury signal by stopping at any sign of hepatotoxicity. Held to that standard, it’s one of the better-evidenced adaptogens on the market — useful for stress, not a hormone therapy.
This article is general educational information, not medical advice. Ashwagandha can interact with sedatives, thyroid, and immunosuppressant medications, is traditionally avoided in pregnancy, and has documented rare liver-injury case reports. Talk to a licensed clinician before starting it, especially if you have liver disease, take prescription medications, or are pregnant or breastfeeding.