Valerian Root for Sleep and Anxiety: What the Evidence Shows

Valerian is the herbal sleep aid that’s been on pharmacy shelves for a century — marketed for racing minds, restless nights, and low-grade anxiety.

Here’s the honest version. Valerian has a real, repeatable signal on how well people say they slept — but that signal mostly evaporates the moment you measure sleep objectively. Across three separate meta-analyses, people taking valerian are more likely to report better sleep than people on placebo, yet the same reviews find essentially no change in how long it actually takes them to fall asleep. It’s cheap, well tolerated, and plausibly useful as a mild, subjective nudge — but the case that it does anything measurable to your sleep is thin, and every trial tests a different, non-standardized extract.

Does valerian actually help you sleep?

On the softest measure — self-reported sleep quality — yes, modestly. On the hard measures, not really.

The most-cited analysis is a 2006 systematic review and meta-analysis (Bent et al.) in The American Journal of Medicine. It pooled 16 randomized, placebo-controlled trials covering 1,093 patients. Six of those trials reported a simple yes/no outcome — did sleep quality improve? — and valerian won: a relative risk of improved sleep of 1.8 (95% CI 1.2–2.9). That sounds encouraging until you read the authors’ two caveats in the same breath: most studies had significant methodological problems, and they found evidence of publication bias in that very summary measure — meaning negative trials were likely missing, inflating the effect.

What happens when you measure sleep objectively?

This is where the story changes, and it’s the part supplement marketing skips.

A 2010 meta-analysis (Fernández-San-Martín et al.) in Sleep Medicine pooled 18 randomized placebo-controlled trials and split the outcomes apart. On the subjective yes/no question, valerian again edged out placebo: a relative risk of 1.37 (95% CI 1.05–1.78) for reported sleep-quality improvement. But on the objective, quantitative measures, the effect collapsed. The mean difference in sleep latency — the time to actually fall asleep — was 0.70 minutes (95% CI −3.44 to 4.83) versus placebo. That’s not a typo: about 42 seconds, with a confidence interval straddling zero. The standardized effect on rated sleep quality was −0.02 (95% CI −0.35 to 0.31) — statistically indistinguishable from nothing. The authors’ conclusion was blunt: valerian may improve subjective insomnia, but its effectiveness “has not been demonstrated with quantitative or objective measurements.”

Put plainly: the strongest evidence for valerian is that it changes how people feel about their sleep more than it changes their sleep. That’s not worthless — the felt experience of insomnia is the thing people actually want relief from — but it’s a very different claim from “it makes you fall asleep faster.”

Does it work for anxiety?

The anxiety evidence is thinner still, and it mostly rides along with the sleep research rather than standing on its own.

The largest synthesis is a 2020 systematic review and meta-analysis (Shinjyo, Waddell & Green) in the Journal of Evidence-Based Integrative Medicine, which reviewed 60 studies (6,894 participants total). Within that, only 8 trials (535 participants) were poolable for an anxiety outcome. The reviewers reported a benefit and concluded valerian “could be a safe and effective herb to promote sleep and prevent associated disorders” — but they were candid that the trials are small, mixed in quality, and inconsistent. Eight modest trials is a slim foundation for an anxiety claim, and none of the major anxiety-disorder treatment guidelines rank valerian as an evidence-based option. If anxiety is your primary problem rather than sleep, the broader anxiety-supplement literature is worth reading with the same skepticism.

Why do the trials keep contradicting each other?

Because they’re not really testing the same thing — and this is the information-gain point most articles miss.

The Shinjyo review’s central explanation for the inconsistent results is the variable quality of the herbal extract itself. “Valerian” isn’t one standardized drug. Trials have used 70% ethanol extracts, aqueous extracts, methanol extracts, whole dried root, and proprietary branded preparations — at doses from 400 to 600 mg and standardized (when at all) to different amounts of valerenic acid. The review notes it’s still uncertain which constituents actually drive any effect, and that some of valerian’s active compounds are chemically unstable, degrading on the shelf. So a positive result for one 0.8%-valerenic-acid extract doesn’t transfer to the capsule you bought — which may contain a different species, a different extraction, and an unknown amount of active compound by the time you swallow it. That heterogeneity is a plausible reason the trials disagree, and it’s a reason to distrust any single glowing study.

So should you take it?

The safety case is genuinely reassuring. Across the 60 studies in the 2020 review, spanning subjects aged 7 to 80, there were no severe adverse events attributable to valerian — the typical complaints are mild dizziness, daytime drowsiness, or stomach upset. That’s a low-risk profile for a short experiment.

Just size your expectations to the data. The realistic upside is a mild, mostly subjective improvement in how restful your nights feel — not a reliable, measurable shortening of the time it takes to fall asleep. If you try it, buy a product that lists its standardization, take it consistently for a couple of weeks, and judge it honestly. And don’t skip the higher-yield levers: the fundamentals behind sleep have far stronger evidence than any single herb, whether that’s magnesium’s role in sleep and anxiety or the tighter, better-studied case for melatonin and circadian timing.

For the bigger picture on non-drug tools for a dysregulated nervous system, see the pillar guide on anxiety regulation and sleep restoration.