Lavender — A Good Result, and Only One Company Has Tested It

My starting point

I don’t reach for capsules when someone has anxiety. I start with what’s feeding it: caffeine, alcohol, sleep debt, blood sugar swings, the workload someone actually carries. Those move anxiety more reliably than anything in a bottle, and they don’t come with a monthly bill. What I see often is the swap: someone wary of a benzodiazepine prescription, or trying to come off one, looking for the same quieting effect with a natural label. Lavender oil capsules are marketed at exactly that person, so I came to this expecting industry gloss over thin evidence. The trials turned out better than I expected, with a catch about who ran them that’s worth knowing before you spend money.

Two very different products

“Lavender” covers two things with very different evidence:

1. A capsule you swallow. Almost all the good trials used one product: Silexan, a standardised lavender oil made by Dr Willmar Schwabe in Germany (sold there as Lasea), taken as 80 mg once a day.
2. Lavender you smell or rub on: aromatherapy, pillow sprays, massage oil. There are many trials, mostly small. Because you can smell the lavender, you know which group you’re in, so most of them can’t be blinded.

Below, each one is graded separately.

The capsule: does it help anxiety?

Grade: mixed, leaning positive. The results are consistent, but all from one company and never independently repeated.

• Five placebo-controlled trials, 10 weeks each, 80 mg a day. They covered anxiety that doesn’t meet a diagnosis, mixed anxiety and depression, and generalised anxiety disorder. Pooled, 1,172 people were analysed. Anxiety scores fell 2.9 points more on Silexan than on placebo, on a 56-point clinician-rated scale. 52% responded on Silexan and 39% on placebo, so about 8 people need to take it for one extra person to respond (Dold et al., 2023).
• Against medicines: in a trial of 539 adults with generalised anxiety disorder, both 80 mg and 160 mg beat placebo. Paroxetine, an SSRI antidepressant, showed only a trend in the main analysis (Kasper et al., 2014). In a six-week trial of people with generalised anxiety disorder, Silexan did as well as a low dose of lorazepam, a benzodiazepine (0.5 mg) (Woelk & Schläfke, 2010).
• Placebo still did a lot. In the largest trial, anxiety scores fell by 9.5 points on placebo, against 12.8 points on 80 mg (Kasper et al., 2014). Most of the improvement people experience would have happened anyway.
• Other reviewers agree on the direction. A network meta-analysis of 29 trials across 12 herbs found a clear effect on anxiety for Silexan (Zhang et al., 2022). But those summaries are pooling the same company’s trials. They can’t make the evidence more independent than it is.

Who paid?

The meta-analysis says it plainly: all the trials “were initiated by the manufacturer of Silexan and performed in one country (Germany)”, and the authors searched the trial registers and found no independent ones (Dold et al., 2023). Schwabe employees are authors on the trials and the pooled analyses, which used individual patient data supplied by the company.

The pattern continues. A 2024 trial in mild-to-moderate depression (498 people) found Silexan beat placebo by about 2 points on a 60-point depression scale, similar to the antidepressant sertraline. Like every Silexan trial before it, it was sponsored by Schwabe (Kasper et al., 2024; Roebuck et al., 2026).

The first independent trial is under way. The STOP trial, run by Deakin University and Phoenix Australia, is testing Silexan 160 mg a day against placebo for 12 weeks as an add-on treatment for PTSD, in a planned 278 adults. Its authors describe it as “the first clinical trial of Silexan conducted by a research team that is independent of the company”. It was registered in May 2024 and its protocol was published in 2026. There are no results yet (Roebuck et al., 2026). It’s a PTSD trial, so even when it reports, it won’t directly repeat the anxiety results. But it will be the first time anyone outside the company has tested this extract blind.

None of that makes the results wrong. The trials are well designed: double-blind, and the placebo capsules were flavoured with a trace of lavender so people couldn’t tell by the smell. But I’d say the same about a drug: one company’s trials, never repeated by anyone else, is a reason for caution.

The capsule: does it help sleep?

Grade: mixed, and only for anxiety-related sleep problems. It isn’t a sleeping pill.

Sleep improved in the anxiety trials. In one trial of 221 people with anxiety and poor sleep, sleep quality scores improved by 5.5 points on Silexan against 3.8 on placebo (Kasper et al., 2010).

But the manufacturer’s own analysis asked why sleep improved. It found that more than 98% of the sleep effect came through reduced anxiety, and almost none directly (Seifritz et al., 2019; Seifritz et al., 2022). It also doesn’t sedate: in a driving-simulator trial, 80 mg was equivalent to placebo (Möller et al., 2020).

So if worry keeps you awake, the capsule has a case. If you’re not anxious and just sleep badly, no trial has tested it for you.

Pillow sprays, oils and baths

Grade: mixed, low quality. Positive on average, with no reliable way to separate the effect from expectation.

There are many meta-analyses of lavender aromatherapy for sleep. Nearly all find an improvement:

• 11 trials, 628 adults: a moderate improvement in sleep quality (Shen et al., 2025).
• 10 trials in people over 60, all inhaled: a larger improvement, with very high variation between trials (I² 87.5%) (Ostovar et al., 2025).
• Across aromatherapy for menopause, the certainty for sleep was graded “very low” (Bagheri et al., 2026).

The problem is built into the method. You can’t hide the smell of lavender, so most trials compare lavender with nothing, or with a scent people can tell apart. Sleep is measured by questionnaire. That’s the situation where expectation does most. A pleasant bedtime ritual may well help. That’s worth having, but the trials can’t tell you whether lavender itself does anything.

Smell is a different route from a capsule, and it’s personal. If you like the scent and it’s part of how you wind down, that’s a reasonable choice. Just don’t expect the capsule trials to tell you anything about the spray.

How it’s meant to work

Grade: mechanistic and untested as the explanation

In lab studies, Silexan dampens calcium channels in nerve cells. In a small brain-scan study of 17 healthy men, eight weeks of 160 mg a day changed serotonin-1A receptor binding (Baldinger et al., 2015). Both are plausible routes to an effect on anxiety. Neither has been shown to explain the trial results.

Safety and who shouldn’t

Grade: safe in trials at the tested dose. One real concern, and it’s about children and the skin, not capsules.

• Side effects in the trials: no more adverse events than placebo overall. The ones attributed to Silexan were mainly mild stomach effects such as nausea and burping (often lavender-tasting) (Dold et al., 2023).
• No abuse potential and no sedation at 80 mg, in the manufacturer’s trials (Seifritz et al., 2021; Möller et al., 2020).
• Other medicines: in 16 healthy volunteers, 160 mg a day for 11 days had no relevant effect on five main liver enzymes that clear drugs (Doroshyenko et al., 2013). Still check with a pharmacist if you take sedatives, antidepressants or anything for anxiety. Don’t swap a prescribed medicine for lavender without your GP.
• Children, and lavender on the skin: in 2007, three boys developed breast growth while lavender and tea tree oil products were being put on their skin. It went away when the products stopped, and both oils had oestrogen-like activity in cell studies (Henley et al., 2007). A 2019 report added three girls and a boy with early breast development linked to continuous use of lavender-scented products (Ramsey et al., 2019). These are case reports, so cause isn’t proven, and the authors say it’s unknown whether the doses involved are enough. It’s reasonable to stop regular lavender oil products on a child’s skin if this happens, and to tell the GP.
• Under 18, pregnancy and breastfeeding: the UK capsule label says not to use it (label below).
• Don’t swallow essential oil from a bottle. The trials used a pharmaceutical-grade capsule with a fixed dose. Aromatherapy oils are sold for scent and the skin.

What to do instead, or first

Grade: strong

For generalised anxiety, psychological therapy has a larger effect than medicines in trials (effect size 0.76 against 0.38). The authors warn that the two can’t be compared directly, because medicine trials use placebo controls (Carl et al., 2020). It’s available through the NHS; ask your GP. For sleep without anxiety, the first-line treatment remains CBT for insomnia, as in last month’s valerian piece.

And the Test, Don’t Guess question: what is the anxiety attached to? Lavender doesn’t answer that. Sleep, caffeine, alcohol, blood sugar swings, thyroid, a life situation: a two-week diary of mood, sleep and what you ate and drank often shows more than any capsule.

How much: what the trials used

• Tested dose: Silexan 80 mg once a day (one trial also used 160 mg), for 6–10 weeks, in adults aged 18–65 with anxiety (Dold et al., 2023; Kasper et al., 2014). In the UK, Kalms One-A-Day contains 80 mg of lavender oil (Lavandula angustifolia) per capsule. It’s a traditional herbal medicine for “the temporary relief of the symptoms of mild anxiety such as stress and nervousness”, “based on traditional use only”. Its leaflet names Dr Willmar Schwabe, the company that makes Silexan, as registration holder and manufacturer, but it doesn’t name the extract, so it doesn’t say whether the oil is Silexan (patient leaflet, THR 05332/0008).
• Upper limit: no official upper limit. Long-term safety data beyond the trials are thin.
• How long, and when to stop: the trials ran 6–10 weeks. If nothing has changed after that, stop.
• Check first if: you’re under 18, pregnant or breastfeeding (the label says don’t use), or you take medicines for anxiety, depression or sleep.
• Grade: mixed, leaning positive for anxiety (one manufacturer); mixed for sleep, and only through anxiety.

The bottom line

• Does it work? For anxiety, probably a little: about one extra responder in eight, all from one company’s trials. For sleep, only if worry is what’s keeping you awake.
• Pillow sprays? Pleasant, harmless for most adults. The trials can’t separate lavender from expectation.
• Is it safe? In adults at 80 mg, it appears to be. Expect some burping. Keep regular lavender oil products off children’s skin if anything unusual happens.
• Better first steps: find what the anxiety is attached to. Talking therapy has the stronger evidence.

This month’s supplement is glycine, tested for sleep to the same standard. This month’s recipe is the nightcap without the alcohol.

Sources

PubMed records resolved and abstracts read 28 October 2026; every PMID and DOI re-checked against PubMed 8 November 2026.

Bagheri L, Ghiasi A, Biabani F. Effects of aromatherapy on menopausal symptoms: a systematic review and meta-analysis. J Obstet Gynaecol 2026;46(1):2723303. PMID 42709539. doi:10.1080/01443615.2026.2723303

Baldinger P, et al. Effects of Silexan on the serotonin-1A receptor and microstructure of the human brain: a randomized, placebo-controlled, double-blind, cross-over study with molecular and structural neuroimaging. Int J Neuropsychopharmacol 2015;18(4) (online 2014). PMID 25522403. doi:10.1093/ijnp/pyu063

Carl E, et al. Psychological and pharmacological treatments for generalized anxiety disorder (GAD): a meta-analysis of randomized controlled trials. Cogn Behav Ther 2020;49(1):1–21 (online 2019). PMID 30760112. doi:10.1080/16506073.2018.1560358

Dold M, et al. Efficacy of Silexan in patients with anxiety disorders: a meta-analysis of randomized, placebo-controlled trials. Eur Arch Psychiatry Clin Neurosci 2023;273(7):1615–1628. PMID 36717399. doi:10.1007/s00406-022-01547-w

Doroshyenko O, et al. Drug cocktail interaction study on the effect of the orally administered lavender oil preparation silexan on cytochrome P450 enzymes in healthy volunteers. Drug Metab Dispos 2013;41(5):987–993. PMID 23401474. doi:10.1124/dmd.112.050203

Henley DV, Lipson N, Korach KS, Bloch CA. Prepubertal gynecomastia linked to lavender and tea tree oils. N Engl J Med 2007;356(5):479–485. PMID 17267908. doi:10.1056/NEJMoa064725

Kalms Lavender One-A-Day Capsules (lavender oil 80 mg), patient information leaflet. Registration holder and manufacturer: Dr. Willmar Schwabe GmbH & Co. KG. THR 05332/0008. MHRA products database, read 8 November 2026. Leaflet (PDF)

Kalms One-A-Day lavender oil 80 mg capsules, product listing and label text, Rowlands Pharmacy (shop.rowlandspharmacy.co.uk), read 28 October 2026.

Kasper S, et al. Lavender oil preparation Silexan is effective in generalized anxiety disorder — a randomized, double-blind comparison to placebo and paroxetine. Int J Neuropsychopharmacol 2014;17(6):859–869. PMID 24456909. doi:10.1017/S1461145714000017

Kasper S, et al. Lavender oil preparation Silexan is effective in mild-to-moderate major depression: a randomized, placebo- and reference-controlled trial. Eur Arch Psychiatry Clin Neurosci 2025;275(6):1693–1703 (online 2024). PMID 38558147. doi:10.1007/s00406-024-01783-2

Kasper S, et al. Silexan, an orally administered Lavandula oil preparation, is effective in the treatment of ‘subsyndromal’ anxiety disorder: a randomized, double-blind, placebo controlled trial. Int Clin Psychopharmacol 2010;25(5):277–287. PMID 20512042. doi:10.1097/YIC.0b013e32833b3242

Möller HJ, et al. Silexan does not affect driving performance after single and multiple dose applications: results from a double-blind, placebo and reference-controlled study in healthy volunteers. J Psychiatr Res 2020;136:543–551. PMID 33221027. doi:10.1016/j.jpsychires.2020.10.028

Ostovar M, et al. A systematic review and meta-analysis of clinical trials on lavender for sleep disorders in older adults. Exp Gerontol 2025;211:112919. PMID 41057090. doi:10.1016/j.exger.2025.112919

Ramsey JT, et al. Lavender products associated with premature thelarche and prepubertal gynecomastia: case reports and endocrine-disrupting chemical activities. J Clin Endocrinol Metab 2019;104(11):5393–5405. PMID 31393563. doi:10.1210/jc.2018-01880

Roebuck G, et al. The Silexan in the Treatment Of Posttraumatic stress disorder (STOP) trial: protocol for a 12-week randomised controlled trial of adjunctive Silexan for PTSD. BMC Complement Med Ther 2026;26(1). PMID 41736003. doi:10.1186/s12906-026-05312-7. ClinicalTrials.gov NCT06412757.

Seifritz E, et al. Effect of anxiolytic drug silexan on sleep — a narrative review. World J Biol Psychiatry 2022;23(7):493–500. PMID 36259937. doi:10.1080/15622975.2021.2013092

Seifritz E, et al. No abuse potential of Silexan in healthy recreational drug users: a randomized controlled trial. Int J Neuropsychopharmacol 2021;24(3):171–180. PMID 33300578. doi:10.1093/ijnp/pyaa064

Seifritz E, Schläfke S, Holsboer-Trachsler E. Beneficial effects of Silexan on sleep are mediated by its anxiolytic effect. J Psychiatr Res 2019;115:69–74. PMID 31121394. doi:10.1016/j.jpsychires.2019.04.013

Shen H, Zhang LJ, Zhu WY. The sleep-enhancing effect of lavender essential oil in adults: a systematic review and meta-analysis. Holist Nurs Pract 2025;40(2):105–118. PMID 40600743. doi:10.1097/HNP.0000000000000734

Woelk H, Schläfke S. A multi-center, double-blind, randomised study of the Lavender oil preparation Silexan in comparison to Lorazepam for generalized anxiety disorder. Phytomedicine 2010;17(2):94–99 (online 2009). PMID 19962288. doi:10.1016/j.phymed.2009.10.006

Zhang W, et al. Medicinal herbs for the treatment of anxiety: a systematic review and network meta-analysis. Pharmacol Res 2022;179:106204. PMID 35378276. doi:10.1016/j.phrs.2022.106204