Ashwagandha for Stress and Anxiety: What the Evidence Shows

Ashwagandha is the supplement aisle’s answer to a wired nervous system — marketed as an ancient “adaptogen” that dials down cortisol and takes the edge off chronic stress.

The honest version is more interesting than the marketing. Ashwagandha has real randomized-trial signals — it lowers cortisol and self-reported stress more than placebo in several controlled studies. But almost every one of those trials is small, short, and run on a single manufacturer’s branded extract, often with that manufacturer’s involvement. And unlike most herbs, ashwagandha carries a documented liver-injury signal that the “natural calm” pitch leaves out. It’s a plausible short-term tool for stress, not a risk-free one.

Does ashwagandha actually lower cortisol?

This is the strongest part of the case. Cortisol is the hormone your HPA axis pumps out under sustained stress, and ashwagandha consistently nudges it down in trials.

In the most-cited study — a 2012 double-blind RCT by Chandrasekhar et al. in the Indian Journal of Psychological Medicine — 64 chronically stressed adults took 300 mg of a standardized root extract (KSM-66) twice daily for 60 days. Serum cortisol fell 27.9% from baseline in the ashwagandha group versus 7.9% on placebo (between-group p = 0.002), and perceived-stress scores dropped 44% versus 5.5%.

A 2019 dose-ranging RCT by Salve et al. in Cureus found the same direction with a cleaner design: 60 stressed adults split across 250 mg/day, 600 mg/day, and placebo for eight weeks. Cortisol dropped significantly at both doses — p < 0.0001 at 600 mg/day — with stress and sleep scores improving alongside. So the cortisol effect isn’t a one-off; it replicates. That’s more than you can say for a lot of supplements sold for anxiety.

Does it reduce anxiety — or is the effect too good to be true?

A 2026 systematic review and dose-response meta-analysis by Alsanie et al. in Complementary Therapies in Medicine pooled 22 randomized trials and reported significant improvements in stress, anxiety, and depression, with a dose-response relationship between ashwagandha and stress reduction (p = 0.031 for the linear trend).

Here’s the information-gain most “benefits” posts skip: the pooled effect sizes were implausibly large — a standardized mean difference of −6.87 for anxiety and −5.88 for stress. For context, an SMD above 0.8 is already considered a “large” effect. Numbers that extreme aren’t a sign of a miracle herb; they’re a red flag for heavy heterogeneity, small-study bias, and inconsistent measurement across trials. The authors say so plainly, concluding that “well-designed, high-quality trials are still needed” before any clinical recommendation. Read the meta-analysis as direction confirmed, magnitude not trustworthy — not as proof of a potent anxiolytic.

How good is the evidence, really?

Three caveats decide how much weight to give all this.

The trials are small and short. The headline studies enroll 60 to 64 people and run 60 days to eight weeks. None tell you what daily use does over a year.

The extract isn’t generic. Most positive trials — including Chandrasekhar 2012 — used KSM-66, a specific branded, standardized root extract (in that study, supplied by its manufacturer, Ixoreal Biomed). A capsule on a shelf that just says “ashwagandha 500 mg” may be a different plant part, a different withanolide concentration, or adulterated entirely. The evidence is for particular standardized extracts, not the category.

Industry fingerprints are everywhere. When a supplement maker funds or supplies the trial of its own product, effect sizes tend to inflate. That doesn’t make the findings fake — but it’s exactly why the implausibly large pooled numbers deserve skepticism rather than a screenshot.

Is ashwagandha safe? The liver-injury signal

For most healthy adults at studied doses, short-term tolerability looks reasonable. A 2026 systematic review by Coope et al. in Pharmaceuticals examined 23 trials totaling 2,317 participants on standardized root extract (125–600 mg/day, up to 180 days) and found hepatic, renal, hormonal, and cardiovascular biomarkers stayed within normal clinical ranges, with no serious adverse events and only mild, placebo-comparable complaints like GI upset, headache, and drowsiness.

But trials screen healthy volunteers and rarely run past six months — and that’s where the real-world signal lives. The NIH’s LiverTox database documents roughly 23 published cases of ashwagandha-linked liver injury, typically appearing 2 to 12 weeks after starting it, usually a cholestatic or mixed pattern with prolonged jaundice. Most resolve within 1–4 months of stopping, but rare cases have required liver transplantation or proved fatal, particularly with pre-existing liver disease. LiverTox rates ashwagandha a “likely cause” of clinically apparent liver injury. Many of those cases involved non-standardized or multi-ingredient products — which loops straight back to the quality problem above. This is genuinely rare, but it’s not nothing, and it’s the part a TikTok rarely mentions.

Who might it help, and how to use it?

If you’re dealing with chronic, grinding stress and want to trial a standardized extract for a defined window — say 8 to 12 weeks at the doses studied (300–600 mg/day of a root-only, standardized product) — the evidence is strong enough to be worth a careful experiment, not a leap of faith. Skip it if you have liver disease, take hepatotoxic medications, are pregnant, or have thyroid or hormone-sensitive conditions, and stop immediately if you notice fatigue, dark urine, or yellowing skin.

What ashwagandha won’t fix is an arousal problem masquerading as a chemistry problem. If your stress is driven by a racing mind and an over-revved nervous system, the lever is regulation, not a capsule — the same reason L-theanine and most calm-in-a-pill products underdeliver against the hype.

The takeaway

Ashwagandha is one of the better-evidenced stress supplements: it reliably lowers cortisol and self-reported stress versus placebo in controlled trials. But the trials are small, short, and tangled with industry funding, the effect sizes are likely overstated, the benefits attach to specific standardized extracts rather than whatever’s cheapest, and a real — if rare — liver-injury risk sets it apart from genuinely inert options. Treat it as a time-limited experiment with a quality product, not a daily habit. For the system this actually sits inside — how to wind an over-aroused nervous system down — see our anxiety regulation and sleep restoration work.