People start ashwagandha on a Monday and want to know by Friday. Fair enough. The trouble is that most of what is written about timing comes from forum posts and brand pages, and the two disagree with each other and with the studies. So here is the question answered the slow way: what the randomised trials measured, at which week, at what dose, in whom. Then a note on our own formula, which uses a lower dose than most of those trials, and why.
Short answer
Most trials measured their main outcome at eight weeks, and most of the significant changes appear at that point. A few trials took interim measurements and found stress scores moving by week four. Nothing in the controlled literature shows a reliable effect in the first week. If someone feels calmer on day three, that is either placebo, the glycine in their evening magnesium, or a very good week. Give it eight weeks and judge it on something you can measure.
What the trials found, and when
Chandrasekhar 2012, 60 days, 600 mg a day. Sixty-four adults with a history of chronic stress took 300 mg of a full-spectrum root extract (KSM-66®) twice daily or placebo. At day 60, serum cortisol had fallen 27.9 % in the ashwagandha group against 7.9 % on placebo, and Perceived Stress Scale scores dropped by about 44 %. Measurements were taken at day 0 and day 60 only, so the trial cannot say whether the change was there at week four (PMID 23439798).
Salve 2019, 8 weeks, 250 or 600 mg a day. Sixty stressed adults were split across placebo, 250 mg and 600 mg of the same extract. Both doses lowered stress scores and serum cortisol at eight weeks; the 600 mg arm did more, including on sleep quality (PMID 32021735). This is the trial that matters for anyone on a lower dose: 250 mg did something, and more did more.
Lopresti 2019, 8 weeks, 240 mg a day. Sixty adults with mild anxiety took 240 mg of a different high-withanolide extract (Shoden) or placebo. Anxiety scores fell more on ashwagandha, with morning cortisol and DHEA-S changes in the same direction. Outcomes were assessed at weeks 4 and 8, with the gap opening by week 8 (PMID 31517876).
Langade 2021, 8 weeks, 600 mg a day. Eighty healthy adults and adults with poor sleep took 300 mg twice daily or placebo. At eight weeks, sleep onset latency, total sleep time and sleep efficiency all improved against placebo, in both groups (PMID 32818573).
Wankhede 2015, 8 weeks, 600 mg a day. Fifty-seven young men on a resistance programme took 300 mg twice daily or placebo. At eight weeks the ashwagandha group had larger strength gains, more arm and chest muscle, and lower exercise-induced muscle damage markers (PMID 26609282).
The pattern across five trials is consistent: doses of 240 to 600 mg of a standardised root extract, outcomes at eight weeks, and where interim data exist, partial movement by week four.
Week-by-week: what to expect
| Week | What the trials show | What people commonly report |
|---|---|---|
| 1 | No controlled evidence of change | "Nothing yet" or a placebo lift |
| 2 | No measured outcomes in the main trials | Slightly easier evenings for some |
| 4 | Stress and anxiety scores begin to separate from placebo (Lopresti) | Less reactive to small stressors |
| 8 | Main effects: cortisol, stress scores, sleep onset, strength (all trials above) | Steadier mood, falling asleep faster |
| 12 | Some trials extend here; effects maintained | Point to reassess whether to continue |
Why it takes weeks, not days
Ashwagandha's active compounds, the withanolides, do not act like caffeine or a sedative. The proposed mechanism is a gradual dampening of the stress axis, the loop between hypothalamus, pituitary and adrenal glands that sets cortisol output. That axis adapts slowly. The trials that measured cortisol saw it fall over two months, not two days. Sleep changes follow the same logic: sleep onset improved at eight weeks, without any acute drowsiness reported on the day of dosing. The NIH's complementary health centre summarises the evidence as promising for stress and sleep, with the caveat that most studies are short and small (NCCIH).
Dose: what "clinically studied" actually means
Every brand says its dose is clinically studied. The trials above used 600 mg a day of KSM-66 (as 300 mg twice), 250 mg of the same extract in one arm, and 240 mg of Shoden. So a serving of 250 to 300 mg is inside the studied range and below the most-cited dose.
Our Cortisol Health for Women carries 300 mg of KSM-66® per two-capsule morning serving. That is half the Chandrasekhar and Langade dose and slightly above the Salve low arm. We chose it because the formula combines ashwagandha with 200 mg of L-theanine, rhodiola, phosphatidylserine and magnesium bisglycinate for daytime use, and because 600 mg in a morning formula made some testers drowsy. How ashwagandha compares with the other three on that list is a separate question, and we answered it in Adaptogens for Stress. Expect the lower arm's result, not the higher one, and expect it at eight weeks. That comparison is on the product page, next to the trial it comes from, and the method for checking any label is in How to Read a Supplement Label.
Morning or night?
The trials dosed twice a day, morning and evening, so the literature does not settle this. Practically: if the goal is daytime stress and focus, morning with food. If the goal is sleep onset, evening. A minority of people feel mildly drowsy in the first week; they belong in the evening camp. Ours is a morning formula because of the L-theanine and rhodiola in it. For the evening we point people at magnesium and at the routine in Why Do I Wake Up at 3 a.m.?.
How to know whether it is working
Feelings drift. Pick two measures on day one and log them weekly:
- A stress score. The Perceived Stress Scale is ten questions, free, and what the trials used.
- Time to fall asleep, estimated each morning.
- Resting heart rate on waking, if you wear a tracker.
- For lifters: a working weight on one lift.
If nothing has moved by week eight on any of them, the honest conclusion is that ashwagandha is not doing much for you, and money spent on month three is money spent on hope.
Side effects and cautions
Most people tolerate it. Mild stomach upset and drowsiness are the common complaints. Three cautions carry more weight:
- Thyroid. Ashwagandha can raise thyroid hormone levels. Anyone with a thyroid condition or on thyroid medication should ask before starting.
- Liver. Rare cases of liver injury have been reported and catalogued by LiverTox, mostly with prolonged use or products of unknown purity. Yellow skin or eyes, dark urine or persistent nausea mean stop and get checked (LiverTox).
- Pregnancy and breastfeeding. Not recommended. Several European food-safety agencies advise against it, and the trials excluded pregnant women.
Also worth a conversation with a prescriber: sedatives, blood-pressure and blood-sugar drugs, immunosuppressants. Ashwagandha can add to their effects.
Should you cycle it?
There is no trial of cycling. The longest well-run studies ran twelve weeks continuously and reported no accumulating harm. A sensible practice, given the thin long-term data, is to reassess at twelve weeks: if the measures you logged have improved and stayed, take a two-week break and see whether they hold. If they hold, you may not need it. If they slide, resume.
Frequently asked questions
How long does ashwagandha take to work for anxiety?
Trials on anxiety and stress measured their main effect at eight weeks, with early separation from placebo around week four.
How long does ashwagandha take to work for sleep?
Eight weeks in the largest sleep trial, at 600 mg a day. It did not act as a same-night sedative.
Can I feel ashwagandha the first day?
Controlled trials show no first-day effect. Some people report mild drowsiness in the first week, which is a timing signal, not a benefit.
What is the best dose of ashwagandha?
240 to 600 mg a day of a standardised root extract, depending on the extract. 600 mg produced the largest cortisol and sleep changes in the KSM-66 trials.
Does ashwagandha work for everyone?
No. In every trial, a share of the treatment group did not improve. Log two measures for eight weeks and let the log decide.
Can I take ashwagandha with magnesium or L-theanine?
Yes; our morning formula combines all three. The trial evidence is for each ingredient separately, not for the combination.
Conclusion
Ashwagandha works on a two-month clock. The controlled trials show cortisol, stress scores, sleep onset and even strength changing at eight weeks on 240 to 600 mg of a standardised root extract, with partial movement by week four and nothing reliable in week one. Buy a product whose dose you can see on the panel, take it at the same time every day, log two measures, and give it eight weeks. Our formula uses 300 mg in a morning stack; the product page shows the trial it is compared against, and the label shows every milligram so you can make that comparison yourself.
References
- Chandrasekhar K et al. Indian J Psychol Med 2012;34(3):255-62. PMID: 23439798
- Salve J et al. Cureus 2019;11(12):e6466. PMID: 32021735
- Lopresti AL et al. Medicine (Baltimore) 2019;98(37):e17186. PMID: 31517876
- Langade D et al. J Ethnopharmacol 2021;264:113276. PMID: 32818573
- Wankhede S et al. J Int Soc Sports Nutr 2015;12:43. PMID: 26609282
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
