Exercise for Insomnia: What the 2024–2025 Evidence Actually Shows

Bottom line: Regular exercise is one of the better-supported non-drug strategies for improving sleep, and recent insomnia-specific reviews find meaningful average benefits. But "exercise helps sleep" is still too broad: subjective improvements are often larger than objective changes, studies vary substantially, and exercise should not be presented as a guaranteed substitute for CBT-I in chronic insomnia.

The modern evidence is encouraging

A 2024 systematic review and meta-analysis included 19 studies in people diagnosed with insomnia or experiencing insomnia symptoms.[1]

Exercise significantly improved both subjective and objective sleep outcomes. The pooled effect was larger for subjective measures, while heterogeneity was high — meaning the size of benefit differed considerably among studies.[1]

That combination is important. Exercise is not just wellness folklore, but the evidence is not precise enough to promise a fixed number of minutes of extra sleep for every person.

Subjective and objective sleep are different outcomes

Someone can report that sleep feels better without showing an equally dramatic change on actigraphy or polysomnography.

That is not automatically a placebo or meaningless result. Insomnia itself is partly defined by the experienced sleep problem and its daytime consequences.

Still, articles should keep the endpoints separate:

• subjective sleep quality asks how sleep felt;
• Insomnia Severity Index measures perceived insomnia burden;
• sleep-onset latency estimates how long falling asleep takes;
• WASO measures wakefulness after sleep begins;
• sleep efficiency describes the fraction of time in bed spent asleep;
• objective measures use tools such as actigraphy or polysomnography.

An intervention that improves one should not automatically be described as improving every dimension.

Is one type of exercise clearly best?

A 2025 network meta-analysis compared multiple exercise approaches across randomized trials, including yoga, Tai Chi, walking or jogging, aerobic plus strength exercise, strength training and mixed aerobic programs.[2]

The interesting result is not a permanent winner. It is that multiple forms of movement appear capable of helping sleep.

Network rankings can shift as new trials arrive and may be sensitive to small study counts. For a real person, adherence is likely to matter: an exercise plan that looks optimal in a ranking but is abandoned after four days has little practical value.

Walking is underrated

Walking has several advantages for sleep experimentation:

• low barrier to entry;
• easy to dose consistently;
• little equipment;
• relatively low recovery cost; and
• it can also increase daylight exposure when done outside.

That last point matters because daytime light can reinforce circadian timing independently of exercise.

Does evening exercise ruin sleep?

The blanket rule "never exercise at night" is too simplistic.

Some people feel activated after intense late sessions. Others sleep normally or better. Relevant variables include:

• exercise intensity;
• session duration;
• how close it ends to bedtime;
• body-temperature recovery;
• stimulants used around training;
• and the individual's normal schedule.

A useful approach is to treat timing as a personal variable to test, not as a universal prohibition.

If high-intensity training within an hour of bed repeatedly delays sleep onset, moving it earlier makes sense. If an evening walk reliably reduces stress without delaying sleep, there is no evidence-based reason to ban it solely because it happens after dinner.

Exercise versus CBT-I

Exercise and CBT-I solve overlapping but different problems.

CBT-I directly targets the mechanisms that maintain chronic insomnia: conditioned wakefulness, excessive time in bed, maladaptive beliefs and unstable sleep behavior. Major guidelines support it as first-line treatment.[3]

Exercise may improve sleep pressure, mood, stress regulation, physical health and circadian stability. Those are valuable, but they do not make stimulus control or sleep scheduling unnecessary in someone with entrenched insomnia.

A better model is exercise plus an insomnia-specific strategy, not exercise versus CBT-I.

How to test exercise without fooling yourself

Track a few simple variables for several weeks:

• exercise type;
• approximate intensity;
• session start and end time;
• bedtime and wake time;
• estimated sleep-onset latency;
• awakenings or WASO;
• subjective sleep quality; and
• next-day energy.

Do not change five other sleep interventions at the same time. If exercise, melatonin, magnesium, bedtime, caffeine and room temperature all change together, you will not know what produced the pattern.

Safety context

Someone who is sedentary, pregnant, recovering from illness or injury, or living with significant cardiovascular, respiratory or musculoskeletal disease may need a more individualized exercise plan.

Severe daytime sleepiness also deserves attention before adding activities that create driving or machinery risk.

Bottom line

Exercise deserves a real place in an evidence-based sleep strategy. Recent meta-analyses support improvements in insomnia and sleep quality, with benefits appearing across several exercise types.[1][2]

The edge is to avoid overselling it. Exercise is useful, low-cost and broadly beneficial, but chronic insomnia still deserves insomnia-specific assessment and, when appropriate, CBT-I.

Related reading

• Insomnia Evidence Guide
• Sleep Hygiene vs CBT-I
• Caffeine and Sleep Timing
• Sleep Onset vs Sleep Maintenance