Skip to content
SleepUncovered

Food and drink for better sleep

Updated16 May 2026Read time5 minReviewed bySleepUncovered editorial

Short answer

Limited evidence

Most foods marketed as sleep aids have weak evidence. The real dietary levers for sleep are caffeine timing, alcohol avoidance, meal timing (no large meals within 2h of bed), and avoiding overnight blood sugar crashes. Specific “sleep foods” like cherries or kiwi have small effects at best.

Key points

  • Caffeine timing matters more than caffeine quantity for most adults — see the half-life calculator.
  • Alcohol within 3h of bed reliably fragments sleep. Even one drink has measurable effects on sleep architecture.
  • Heavy meals within 2h of bed delay sleep onset and worsen sleep quality.
  • A protein-and-fat snack before bed stabilises blood sugar for night-wakers; carbs alone can cause rebound crashes.
  • Tart cherry juice, kiwi, walnuts: small-study evidence for sleep onset, but effect sizes are modest.

The four real levers

1. Caffeine timing

See the dedicated caffeine page. Headlines: 5-hour average half-life with huge individual variation; even afternoon caffeine measurably suppresses deep sleep in many people. The caffeine cut-off calculator gives a personalised cut-off.

2. Alcohol

The most reliable sleep-disruptor in most people's diets. Even a single drink within 3h of bed measurably increases wake-after-sleep-onset and suppresses REM. See the alcohol page.

3. Meal timing

Large meals within 2 hours of bed:

  • Delay sleep onset (active digestion and post-prandial alertness).
  • Increase reflux risk, particularly lying down within an hour of eating.
  • Disrupt body temperature regulation needed for sleep.

Earlier dinners — 3+ hours before bed — consistently produce better sleep in studies. The effect is modest but reliable.

4. Overnight blood sugar

If you wake at 2–4am feeling alert, hungry, or with a racing heart, low overnight blood sugar may be the cause. Cortisol rises to mobilise glucose; the cortisol spike breaks sleep.

Fix: a small protein-and-fat snack 30 minutes before bed — e.g. a handful of nuts, Greek yogurt, peanut butter on a cracker. Carbs alone can cause rebound hypoglycaemia and worsen the problem.

The “sleep foods” evidence

Tart cherry juice

Contains low levels of melatonin. Small studies show ~15 min improvement in sleep time with 240ml twice daily. Effect size modest. The melatonin dose from cherry juice is roughly 0.1 mg — much less than a typical supplement.

Kiwi

One small Taiwanese study (Lin et al, 2011) showed two kiwifruits an hour before bed improved sleep onset and quality. Not widely replicated, but plausible — kiwis contain serotonin precursors and antioxidants. Low-cost low-risk experiment.

Walnuts

Contain melatonin and tryptophan. Small mechanistic case, no strong sleep trial evidence in humans.

Warm milk

Tryptophan content is real but small. The classic “warm milk helps sleep” is probably mostly comfort and ritual. Honest assessment: small placebo-amplifying effect, no meaningful direct mechanism.

Chamomile tea

Mild anxiolytic. Possible small sleep-promoting effect at higher concentrations than typical bagged tea. Worth trying; don't expect much.

Bananas

Magnesium and potassium content is real but modest. Marketing claims exceed evidence.

What to actually do

  • Last caffeine 8–10h before bed. Earlier if you're sensitive.
  • No alcohol within 3h of bed if sleep matters that night.
  • Dinner 3h before bed; if later, keep it light.
  • If you wake middle-of-night reliably, try a small protein-and-fat snack 30 min before bed for a week.
  • Don't bother spending money on “sleep teas” or specialty foods marketed for sleep. The hygiene basics matter much more.

Related reading