Ashwagandha for Cortisol: What the Trials Show

Ashwagandha shows up in a lot of “cortisol support” bottles. The pitch is simple: take it, your cortisol comes down, you stop feeling wired. The ashwagandha cortisol research is more interesting than the pitch, and a lot less tidy.

There are real randomized trials here, and several found lower cortisol than placebo. But most are small and short. Most were paid for, or supplied, by the company that makes the extract being tested. They used different extracts at different doses. And the one trial I found in men aged 40 to 70 found no cortisol difference versus placebo.

This article goes through the key trials one at a time, then the meta-analyses that pool them, then sleep, funding and safety. Each study gets its dose, size and funder next to it, or a note where I couldn’t confirm one.

A disclosure, because it matters on this topic: this site earns commissions on CortiSync, a supplement whose label lists Sensoril, one of the extracts below. Read my opinions with that in mind, and check them against the receipts.

How to read an ashwagandha cortisol study

Three things to know before reading any ashwagandha headline.

Cortisol gets measured in different ways. Some trials drew blood (serum or plasma). Others used saliva. Those aren’t interchangeable, and cortisol swings across the day by design. Expert consensus guidelines on the morning cortisol rise stress that meaningful readings depend on tightly controlled sampling. A trial’s cortisol result is only as good as its sampling protocol. For why a single reading is hard to interpret, see what high cortisol really looks like in men.

Stress is usually a questionnaire. A common measure is the Perceived Stress Scale (PSS), a short self-report. Others include the DASS-21 and a clinician-rated scale, the HAM-A. These are stress-questionnaire scores. They don’t always move with cortisol.

Watch for “from baseline.” A drop from the start of the study in the ashwagandha group is not the effect of ashwagandha, because placebo groups improve too. The number that matters is the difference versus placebo.

The key trials, one receipt at a time

KSM-66 at 600 mg/day: Chandrasekhar 2012

This is the trial quoted most often. Chandrasekhar and colleagues randomized 64 adults with a history of chronic stress at a single hospital in India. The active group took 300 mg of a root-only extract twice a day for 60 days. Secondary sources identify the extract as KSM-66.

The ashwagandha group improved more than placebo on three stress questionnaires and had a larger fall in serum cortisol. Side effects were mild and similar in both groups. The figure you’ll often see quoted, a drop of about 27.9% in cortisol, is the change from baseline within the ashwagandha group, as reported by secondary sources. It is not the difference versus placebo.

KSM-66 at two doses: Salve 2019

Salve and colleagues split 60 stressed but otherwise healthy adults in India into three groups: 250 mg/day, 600 mg/day or placebo, for 8 weeks. Both doses were associated with lower PSS scores and lower serum cortisol than placebo, and with better sleep quality. The drop on the clinician-rated HAM-A was mainly at 600 mg/day.

With about 20 people per group, the trial didn’t formally establish that one dose beat the other. The KSM-66 maker reportedly supplied the product.

Shoden in stressed adults: Lopresti 2019

Shoden is a different extract, made from root and leaf. In a 60-day trial of 60 stressed, healthy adults, Lopresti and colleagues found that 240 mg/day was associated with a larger fall in morning cortisol and DHEA-S than placebo, and a significant drop on the HAM-A. The main self-report stress measure, the DASS-21, only came close to significance. Arjuna Natural, which makes Shoden, reportedly funded the project and supplied the extract.

Sensoril: Auddy 2008 and Pandit 2024

Sensoril, which its supplier describes as a root-and-leaf extract, rests on two stress trials that each tested 125, 250 and 500 mg/day. The older one, Auddy 2008, enrolled 130 chronically stressed adults in India for 60 days, and 98 finished. The Sensoril groups were reported to have lower serum cortisol and C-reactive protein, lower stress-questionnaire scores and higher DHEA-S than placebo, with larger effects at higher doses according to secondary summaries.

That trial appeared in a small nutraceutical-trade journal and isn’t indexed in PubMed. Search summaries report that the lead author worked for Natreon, which developed Sensoril, and a co-author for the company marketing it. Read it as a company-funded, manufacturer-linked trial.

The newer one, Pandit 2024, ran for 8 weeks. Sensoril was associated with dose-related falls in PSS scores and in plasma cortisol, ACTH and salivary alpha-amylase compared with placebo. The extract was reportedly prepared by Natreon India, so this one is manufacturer-linked too. The participant counts I’ve seen reported don’t add up, so I’m not quoting one.

Get the 14-Day Fade Reset

Five short emails over two weeks: light, caffeine timing, walking, a sleep window and a stress log. Habits, not pills.

Sign-ups open soon.

The trial closest to you: men aged 40 to 70

If you’re the reader this site is written for, this is the trial to weigh most. Lopresti and colleagues ran a crossover study in overweight men aged 40 to 70 with mild fatigue. Each man spent 8 weeks on Shoden and 8 weeks on placebo. Of 57 men who enrolled, 43 completed all 16 weeks.

Compared with placebo, there was no significant difference in salivary cortisol. There was also no significant difference in fatigue or vigour. Other hormone results were secondary outcomes and aren’t covered here. Arjuna Natural, the Shoden maker, reportedly funded it.

This doesn’t prove ashwagandha does nothing for men in midlife. It’s one small trial, with one extract and saliva-only hormone measures. But it’s the most direct test I could find for men this age, and it didn’t find a cortisol difference.

What the meta-analyses add

Pooling small trials helps, up to a point. Four reviews are worth knowing.

Akhgarjand 2022. Across 12 randomized trials with 1,002 adults, ashwagandha was associated with lower stress scores than placebo, with a standardized mean difference of -1.75. Effects that large usually signal small, varied trials, not a likely real-world result. The authors rated the certainty of evidence as low, with high heterogeneity. Cortisol wasn’t the headline outcome.
Della Porta 2023. A systematic review of 9 trials in stressed adults, which described the trials rather than pooling them, found ashwagandha was associated with lower cortisol and no significant adverse effects. The authors concluded that short-term use appears to reduce stress, but long-term effects aren’t understood and use should be medically supervised.
Arumugam 2024. This meta-analysis found lower PSS scores, lower HAM-A scores and lower serum cortisol compared with placebo. Side effects were limited, but long-term safety data were lacking.
Albalawi 2025. The useful counterweight. Pooling 7 trials on cortisol and 6 on perceived stress (488 participants), this single-author review found a significant fall in cortisol of 1.16 µg/dL on average, but no significant effect on Perceived Stress Scale scores.

Two cautions. First, these reviews draw heavily on the same small pool of trials, so four reviews are not four separate confirmations. Second, a lower cortisol number did not reliably mean people felt less stressed.

My summary: on average, ashwagandha is associated with modestly lower cortisol in short trials of stressed adults. The certainty is low, heterogeneity is high, and whether people feel less stressed varies from review to review.

Sleep: where dose and time seem to matter

Sleep is where the pooled data are a little clearer. A 2021 meta-analysis of 5 trials with 400 adults found a small but significant effect on overall sleep scores. The effect was stronger at 600 mg/day or more and with 8 weeks or more of use. Mental alertness on waking also improved. Quality of life didn’t change, and no serious side effects were reported.

That’s only five trials, with moderate heterogeneity. The stronger effects came at higher doses taken for longer, which is worth checking against any label. If broken sleep is your main problem, start with the habits in why you keep waking up at 3am before you reach for a capsule. If it has gone on for months, raise it with a clinician rather than a supplement shelf.

Who paid for the research

Here’s the funding picture for the trials above, as far as I could confirm it.

Shoden: both Lopresti 2019 trials were reportedly funded by Arjuna Natural, the maker.
Sensoril: both stress trials are manufacturer-linked to Natreon, the developer.
KSM-66: the maker reportedly supplied the product for Salve 2019. The funding statement for Chandrasekhar 2012 wasn’t verified.
Reviews: I couldn’t confirm funding statements for the four reviews or the sleep meta-analysis.

Company money doesn’t make a result wrong. But it does mean most of this evidence comes from makers testing their own extracts, and outside replication is thin. That’s a reason to read effect sizes conservatively.

Not all ashwagandha is the same extract

“Ashwagandha” on a label can mean very different things. KSM-66 is marketed as a root-only extract. Sensoril and Shoden are described by their makers as root-and-leaf extracts. The makers of Sensoril and KSM-66 standardize to different markers, and the doses tested ran from 125 mg/day to 600 mg/day.

The NIH Office of Dietary Supplements says that because trials used different preparations and doses, no specific extract or amount can be recommended. A result for one extract doesn’t automatically carry over to another. The two you’ll see most, and the trials behind each, are compared in Sensoril vs. KSM-66.

Safety: what’s known and what isn’t

For most healthy adults in short trials, ashwagandha looks well tolerated. The gaps are in rare events and long-term use.

Short-term tolerance. The ODS fact sheet describes ashwagandha as generally well tolerated for up to about 3 months, with mostly mild side effects such as loose stools, nausea and drowsiness. It notes a lack of long-term safety data, possible drug interactions, and cautions for pregnancy and hormone-sensitive prostate cancer.
Rare liver injury. The NIH’s LiverTox entry notes that clinical trials haven’t reported liver enzyme rises or clinically apparent liver injury, but rare cases have been reported, typically 2 to 12 weeks after starting, with jaundice and itching. They usually resolve within months. A case series from Iceland and the US describes five such patients; none developed liver failure.
Existing liver disease. A case series from India reported that some patients with pre-existing chronic liver disease developed acute-on-chronic liver failure after using single-ingredient ashwagandha, and some of them died. Product quality couldn’t be ruled out, but it’s a strong reason for extra caution.
Thyroid. In a small trial of adults with subclinical hypothyroidism (a raised TSH), 600 mg/day of root extract for 8 weeks raised T3 and T4 and lowered TSH compared with placebo.

Talk to your clinician first, especially if you’re pregnant, take medication, or have a thyroid, liver or autoimmune condition. If you notice yellowing of your skin or eyes, or itching that won’t settle, stop and contact a clinician promptly.

And if you’re worried your cortisol is truly high for a medical reason, an herb isn’t the answer. That’s a job for a clinician and proper tests, which the high cortisol symptoms guide walks through.

How I’d use this evidence

Opinion · Ray · September 26, 2026

I wouldn’t start with ashwagandha. I’d start with the cheaper levers: a fixed sleep window, a caffeine cut-off, some daylight in the morning and fewer weeknight drinks. The free 14-Day Fade Reset covers most of those, and the Fade Score points you at the habit to fix first.

If I did try it, I’d pick a product that names its extract and prints the dose per serving, and I’d match a dose that was actually studied for that extract. I’d give it 8 weeks, track one thing I can feel, like sleep or a weekly stress score, and skip the home cortisol test. And I’d clear it with a clinician first.

One person's view. Your results may differ.

If you want to see how one Sensoril-based formula stacks up against these trials, that’s in my CortiSync review. It’s an affiliate page, and it says so at the top.

Bottom line

Ashwagandha has a real set of randomized trials behind it, and the pooled data point to modestly lower cortisol on average. That’s the honest good news.

The rest is caveats. The trials are small, short and mostly manufacturer-linked. They tested different extracts at different doses. A lower cortisol number didn’t reliably mean people felt less stressed. And the one trial in men aged 40 to 70 found no cortisol difference. Sleep results look better at 600 mg/day or more over at least 8 weeks.

If you try it, pick a named extract at a studied dose, give it time, track how you feel, and talk to your clinician first, especially with a thyroid or liver condition.

Ashwagandha for Cortisol: What the Trials Actually Show