The short version
Four food categories have real research supporting sleep benefits: tart cherries, kiwi, fatty fish, and magnesium-rich foods. Most popular "sleep foods" - warm milk, chamomile, turkey - have weak or null evidence. The ranked list below is calibrated to actual effect size from published research, not wellness marketing.
Tart cherries (Montmorency variety)
Evidence level: strong
Multiple randomized controlled trials show tart cherry juice increases total sleep time by 25 to 85 minutes and improves sleep efficiency[1]. The mechanism is measurable melatonin content plus inhibition of tryptophan degradation (leaving more serotonin precursor available).
How to use: 8 oz tart cherry juice 1 hour before bed, or 1 to 2 servings of tart (sour) cherries. Most people see benefit within 7 to 14 days of consistent use. Sweet Bing-type cherries do not have the same research - it's specifically the Montmorency variety.
Watch the sugar: tart cherry juice contains 20g+ sugar per serving. Diluting with water helps; tart cherry capsules are an alternative if sugar is a concern.
Kiwi fruit
Evidence level: strong
A well-known study had adults with sleep problems eat 2 kiwis 1 hour before bed for 4 weeks. Sleep onset time dropped 35 percent, total sleep time rose 13 percent, and sleep efficiency rose 5 percent[2]. The likely mechanism is high vitamin C (supports serotonin synthesis) plus actinidin enzymes that modestly affect digestion.
How to use: 2 kiwis 1 hour before bed. Peel and eat as-is, or halve and scoop. Works well for people with late-evening schedules when a heavier snack would disrupt sleep.
Fatty fish
Evidence level: moderate to strong
Regular fatty fish consumption improves sleep quality, likely via vitamin D and long-chain omega-3 contributions to serotonin synthesis and circadian regulation. Studies on salmon intake 3x per week show improved sleep efficiency and faster sleep onset over 6-month periods.
How to use: salmon, sardines, mackerel, herring, or anchovies 2 to 3 times per week. Supplements (2g fish oil with EPA+DHA) are a reasonable fallback if fish is logistically hard. Benefits compound over 6 to 12 weeks.
Magnesium-rich foods
Evidence level: moderate
Magnesium plays a role in regulating GABA (the brain's primary calming neurotransmitter) and supporting melatonin function. Supplementation studies show modest improvements in sleep efficiency and onset, especially in adults with low baseline intake[3]. About 50 percent of Americans have suboptimal intake.
Food sources:
- Pumpkin seeds (150mg per quarter-cup)
- Spinach (1 cup cooked: 160mg)
- Swiss chard (1 cup cooked: 150mg)
- Almonds (1 oz: 80mg)
- Black beans (1 cup cooked: 120mg)
- Dark chocolate 70% (1 oz: 65mg)
- Avocado (1 medium: 60mg)
Target: 350 to 400mg daily from a mix of sources. Supplement form: magnesium glycinate or threonate preferred (better absorption, less GI upset than magnesium oxide).
What has weak or null evidence
Warm milk
Cultural comfort food. The tryptophan content is not uniquely high, and warm milk's effect on sleep in controlled trials is indistinguishable from the placebo effect of a warm beverage. If it helps you, continue - the placebo effect is real, just not pharmacological.
Chamomile tea
Modest, inconsistent evidence. Some small studies show slight sleep quality improvements. Most trials are small and poorly controlled. The anxiolytic effect of any warm relaxing tea before bed likely accounts for most of the benefit.
Valerian root
Popular but evidence is mixed. Systematic reviews find some improvement in subjective sleep quality but not in objective measures. Worth trying individually, not worth betting on.
Turkey at Thanksgiving
No. The "turkey tryptophan" story is folklore. The drowsiness after Thanksgiving dinner is the meal size, alcohol, and afternoon timing. Any large holiday meal produces the same effect.
Lavender (ingested)
Aromatherapy lavender has real evidence for stress and sleep. Eating lavender or drinking lavender tea has thin evidence - most of the documented effect is olfactory.
Foods that hurt sleep
Less sexy but more actionable than the "what to eat" list:
Alcohol within 3 hours of bed
Strong evidence for sleep disruption even at moderate doses. Shortens sleep onset but fragments the second half of the night, especially REM sleep. One glass of wine with dinner is fine. Two drinks within 3 hours of bed measurably hurts sleep.
Heavy meals within 2 hours of bed
Both large portions and late timing affect sleep. Even otherwise "good for sleep" foods become counterproductive if eaten at 10pm before an 11pm bedtime. The full case and 2-week test protocol is in the case for an earlier dinner.
High-caffeine drinks after noon
For slow caffeine metabolizers (10 to 15 percent of adults), this includes afternoon iced coffees, diet sodas, pre-workouts, and chocolate after dinner. The 4-day test for whether morning coffee is wrecking your sleep is the cleanest way to rule caffeine in or out.
High-sugar desserts at night
Glucose spikes and the reactive crash produce micro-wake-ups between 2 and 4am that show up as fragmented sleep on trackers.
How to stack the best foods
A realistic evening routine that incorporates the evidence-backed options:
- 6:30pm: dinner with salmon or sardines 2-3x per week, leafy greens, and a reasonable carb source
- 8:30pm (1 hour before bed): tart cherry juice OR 2 kiwis
- Evening snack if hungry: almonds + a small square of dark chocolate (both contain magnesium)
- Avoid: alcohol after 7pm, caffeine after noon, very sweet desserts
This is not dramatic. The effects are cumulative over weeks. Expect modest but real improvements in deep sleep, sleep onset, and morning alertness within 3 to 4 weeks. A tracker score alone will not make the cause visible - why a sleep tracker cannot explain a bad night covers why you need a parallel input log.
When to see a doctor
If sleep is consistently poor despite good food timing and sleep hygiene, the issue is usually not diet. Sleep apnea, insomnia disorder, anxiety, thyroid dysfunction, and circadian rhythm disorders all require workup beyond dietary changes. A sleep medicine physician or at minimum a primary care visit is the right next step after 4 to 6 weeks of good sleep hygiene without improvement.
