Yes, modestly, on average: a meta-analysis of 66 studies found regular exercise had small effects on how long and how efficiently people slept, and a moderate effect on how well they rated their sleep1. The effect is not a switch, and it is clearest in people whose sleep has room to get better.
This page covers what the studies measured, how big the effects were, whether lifting counts, and who gains most.
What was studied
Sleep research splits exercise into two kinds. A single session (acute exercise) is one workout, followed by that night's sleep. Regular exercise means training over weeks, then comparing sleep before and after, or against a group that did not train.
The main sleep measures, in plain words:
| Measure | What it means |
|---|---|
| Total sleep time | How long you actually slept |
| Time to fall asleep | Minutes from lights out to sleep |
| Sleep efficiency | Share of time in bed actually asleep |
| Wake after sleep onset | Minutes awake after first falling asleep |
| Sleep quality | How well people rate their own sleep, usually on a questionnaire |
The broadest review pooled 66 studies published up to May 2013, covering both single sessions and regular exercise1. It was a meta-analysis, a study that combines many studies on the same question. It graded results by effect size, a standard way to say how big a difference is: small, small-to-medium or moderate.
In plain terms: one workout and one night, or weeks of training and a before-and-after.
What the studies found
After a single session. Small benefits for total sleep time, time to fall asleep and sleep efficiency, and a moderate benefit for time awake during the night1.
After regular exercise. Small benefits for total sleep time and sleep efficiency, small-to-medium for time to fall asleep, and moderate for self-rated sleep quality1.
What changed the size of the effect. Sex, age, how active people were to begin with, type of exercise, time of day, session length and how well people stuck to the plan all mattered1. Exercise intensity did not, and neither did whether the exercise was aerobic or anaerobic, roughly cardio versus short hard efforts like lifting1. We read that review as an abstract, which gives the size categories but not the numbers behind them.
Lifting specifically. A 2018 systematic review of 13 studies on resistance exercise (lifting weights or similar) concluded that regular lifting was linked with better results on every sleep measure studied, most of all sleep quality3. The benefit looked smaller when lifting was combined with cardio and compared with cardio alone. Single lifting sessions were poorly studied and gave inconsistent results3.
A 2025 randomised trial put numbers on it5. 160 older women, average age 69, were assigned to 12 weeks of supervised lifting or to a control group, and split by whether they started as good or poor sleepers. The women were rated with the Pittsburgh Sleep Quality Index, a questionnaire where a lower score means better self-rated sleep. In the poor sleepers who lifted, the score fell from 7.26 to 4.61, ending lower than in both control groups5. The authors concluded that self-rated sleep quality improved regardless of how well the women slept at the start5.
In plain terms: exercise, lifting included, shifted sleep in a good direction on average, by small to moderate amounts.
Who gains most
The size of the effect depends on how much room there is to change. That pattern shows up across the studies.
A 2017 review of three earlier meta-analyses (950 adults) pooled self-rated sleep quality and found a standardised effect of −0.50 (95% CI −0.72 to −0.28), where negative means better sleep2. A confidence interval (CI) is the range the true value most likely sits in. But those studies were in specific groups: adults with obstructive sleep apnoea, adults with cancer, and older adults with sleep problems2. The authors rated the quality of that evidence as very low to low2.
In healthy adults who already sleep well, the picture is less clear. A 2026 meta-analysis of randomised trials found that long-term aerobic training shifted sleep quality in the same direction, but the result was not statistically reliable (−0.53, 95% CI −1.33 to 0.27)6. A 2022 analysis of 28 trials in good sleepers found evening exercise did not disrupt sleep, and the authors noted they found no benefit either, which they put down to ceiling effects: good sleep has little room to get better4.
Self-rated sleep quality with exercise vs no exercise, by population · SMD
In plain terms: the worse the starting sleep, the more there was to gain in these studies.
The nuance
Self-rated versus measured. The largest effects were on sleep quality, which is how people rate their own sleep on a questionnaire15. Measured sleep time and efficiency moved less1. Both matter, but they are not the same thing.
Effect sizes are relative. "Small" and "moderate" describe differences on a standard statistical scale, not minutes of sleep. The 66-study review's abstract does not translate them into minutes, so neither do we.
Evening timing is its own question. Whether a late workout disturbs that night's sleep is covered in the companion article "Does lifting at night ruin your sleep?". In short, the pooled trials in healthy adults found no disruption4.
Association is not always cause. The randomised trials here support cause and effect for the groups studied. Reviews that pool different designs and populations are weaker; the 2017 review flagged very low to low quality evidence2.
In plain terms: expect nudges, most visible in how rested you feel, not a sudden change in the numbers.
What it means in practice
The trials describe group averages. A few descriptive points:
- Regular beat single. The larger effects came from exercise over weeks, not one session1.
- Lifting counts. Type of effort, cardio or lifting-type work, did not significantly change the effect in the 66-study review1, and a lifting trial found clear changes in poor sleepers5.
- Sticking with it mattered. Adherence was one of the factors that changed the effect size1.
For building a weekly routine you can keep, see how many days a week to lift and the beginner workout program. If sleep problems are persistent, a doctor is the right person to ask; nothing here is a sleep plan.
The studies at a glance
| Study | Who was studied | Key result | Year |
|---|---|---|---|
| Kredlow et al., *J Behav Med*1 | Meta-analysis of 66 studies, single sessions and regular exercise (abstract) | Regular exercise: small effects on sleep time and efficiency, moderate on self-rated quality | 2015 |
| Kelley & Kelley, *J Evid Based Med*2 | Review of 3 meta-analyses, 950 adults with sleep apnoea, cancer, or older with sleep problems | Sleep quality effect −0.50 (95% CI −0.72 to −0.28); very low to low quality evidence | 2017 |
| Kovacevic et al., *Sleep Med Rev*3 | Systematic review of 13 resistance-exercise studies (abstract) | Regular lifting linked with better results on all sleep measures; single sessions inconsistent | 2018 |
| Yue et al., *Nat Sci Sleep*4 | 28 crossover trials, 325 healthy good sleepers | Evening exercise did not disrupt sleep; no benefit found, likely a ceiling effect | 2022 |
| Cunha et al., *Psychol Sport Exerc*5 | 160 older women, 12 weeks of supervised lifting vs control (abstract) | Poor sleepers’ quality score fell from 7.26 to 4.61 | 2025 |
| Wang et al., *Front Hum Neurosci*6 | Randomised trials in healthy adults, long-term aerobic training | Sleep quality effect −0.53 (95% CI −1.33 to 0.27), not statistically reliable | 2026 |
Limits worth knowing
- Abstract only for Kredlow 2015, Kovacevic 2018 and Cunha 2025. The 66-study review's abstract gives size categories, not values, so no number is drawn for them.
- Mixed populations. Larger effects came from groups with sleep problems or health conditions; healthy good sleepers showed less.
- Mostly self-rated. Sleep quality is a questionnaire score, open to expectation effects when people know they are exercising.
- Old search dates. The 66-study review covers studies up to 2013.
- Short follow-up. Most training studies ran 12 weeks or less; the 2026 review notes a severe lack of follow-up beyond 6 months6.
- High variation between studies in the 2026 review (I² 90%, a measure of how much results disagree)6.
Latest research
The newest review we opened, published in 2026, pooled 19 randomised trials in healthy adults on aerobic exercise, mood, thinking and sleep6. For sleep, the trials of long-term aerobic training pointed toward better self-rated sleep quality, but the pooled result was not statistically reliable (−0.53, 95% CI −1.33 to 0.27), with high disagreement between studies6. The authors suggest too few participants and ceiling effects as possible reasons. It is a reminder that in people who already sleep well, the average gain may be small.
The short version
- Regular exercise was linked with small gains in sleep time and efficiency and a moderate gain in self-rated sleep quality.
- Cardio versus lifting-type work made no significant difference in the largest review.
- People starting with poor sleep showed the clearest changes; good sleepers had less room to move.
- It is a nudge measured over weeks, not an overnight fix.
Educational overview only — not medical advice.
