Self-rated sleep quality improves a littleEmerging evidence
In 19 trials of 1,832 women with sleep problems, yoga improved sleep quality scores by about 1.2 points on the 21-point PSQI. Every trial had a high or unclear risk of bias, and smaller trials more often found an effect. Across 86 exercise trials (7,276 adults, not chosen for insomnia), yoga improved sleep quality against no exercise but ranked last of six exercise types, with no significant differences between the types; another network of 27 trials ranked Pilates first.
Sources 1, 2, 27
Insomnia severity and measured sleep: little or no changeEmerging evidence
In the same review, three trials found no effect of yoga on insomnia severity, and there was no effect on sleep measured by a wrist device. In 186 midlife women, sleep on wrist devices didn't change with yoga. Across exercise trials in insomnia, self-rated sleep improved more than measured sleep.
Sources 1, 28, 22
Yoga against CBT-I: the YOCAS-II trialEmerging evidence
In about 740 cancer survivors who met the criteria for insomnia, insomnia severity fell 3.61 points with 4 weeks of yoga (75 minutes twice a week), 7.22 with 8 weeks of CBT-I and 2.19 with health education. Yoga was 3.52 points behind CBT-I, well past the margin the trial set for calling them equal, and still 2.78 points behind at 6 months, though it beat health education. Fewer people dropped out of yoga (16% against 30% for CBT-I).
The results come from the trial registry and a conference abstract; no journal paper has appeared, and a 2025 Cochrane review counts it from unpublished data.
Sources 8, 9, 26
Midlife women with hot flushesEmerging evidence
In 249 women, 12 weeks of yoga lowered insomnia severity by about 1.3 points more than usual activity, with no change on wrist devices. Across four trials in 546 women, CBT-I cut insomnia severity by 5.2 points, while yoga's fall was small.
Sources 29, 28, 30
Menopause
NICE suggests menopause-specific CBT for sleep problems that come with hot flushes and night sweats, and the NHS lists yoga, tai chi or meditation among relaxing things to try. Our yoga for menopause guide covers it.
Sources 31, 32
Cancer survivors with poor sleepEmerging evidence
In 410 survivors with disturbed sleep (not diagnosed insomnia), 4 weeks of yoga improved sleep quality by 0.79 points more than usual care, and people in the yoga group cut their sleeping tablet use by 21% a week. Across 25 trials in people with cancer, walking improved sleep problems more than yoga.
Sources 19, 33
Older adultsEmerging evidence
In a network of 35 trials in older adults, yoga wasn't among the exercise types that beat usual care; in 45 studies of sleep problems, yoga improved sleep quality scores, but only meditation beat sleep hygiene education. The older adults' insomnia trial that is often quoted gives no sample size in its summary and doesn't say it was randomised.
Sources 34, 35, 21
The small insomnia trials are weakMixed evidence
Kundalini yoga beat sleep hygiene advice in 44 people, but the authors chose sleep hygiene as a weak comparison, the lead author taught both groups, sleep was self-reported, and they say the effects are "likely overestimates". Other insomnia trials had 44 postmenopausal women across three groups, or no comparison group at all (20 and 21 people who finished).
Sources 20, 36, 37, 38
Against other exerciseEmerging evidence
In 240 women with bladder leaks, 3 months of yoga didn't improve sleep more than time-matched general exercise (a 0.29-point difference). In 226 people with generalised anxiety disorder, sleep improved no more with yoga than with stress education.
Sources 10, 11
The "nearly 2 hours" headlineMixed evidence
A 2026 network of 22 insomnia trials (1,348 people) estimated that yoga added 111 minutes of total sleep against usual care. But it was "all based on sleep diary data", the number of yoga trials behind it isn't given in what we could read, and 68% of the trials had some concerns about bias. Trials that measured sleep with devices found nothing like it.
Sources 39, 1, 28
Exercise in insomnia, and a tai chi trialEmerging evidence
Across 24 trials (1,591 people with insomnia), physical activity improved self-rated sleep, with low certainty. Over 6 months or more, exercise and CBT-I both helped, CBT-I more. In 200 adults over 50, tai chi was worse than CBT-I at 3 months but as good at 15 months. Our searches found no such trial of yoga outside cancer.
Sources 40, 41, 15, 42
Is it safe?Emerging evidence
In the 19-trial review, yoga "was not associated with any serious adverse events", though only 2 trials reported on safety. YOCAS-II listed 3 serious adverse events with yoga and 1 with CBT-I, with no clear link to the programmes. NCCIH (the US National Center for Complementary and Integrative Health) says "Serious injuries are rare."
Sources 1, 8, 43