BackNovember 22, 202611 min readEmmaCentury

Do Sleep Aids Actually Work? What the Evidence Says

Melatonin, antihistamines, magnesium, alcohol, and prescription pills — what each one actually does to your sleep, ranked by evidence. Plus a two-week self-test you can run with the watch on your wrist.

Updated 2026-11-22Editorial standards
Do Sleep Aids Actually Work? What the Evidence Says

It's 2:40am. You've been awake long enough to start negotiating with yourself. You're thinking about the bottle in the bathroom cabinet, the melatonin gummies in the kitchen drawer, or the half-glass of wine that might just take the edge off.

The honest answer about sleep aids is messier than the marketing: some of them work for specific problems, most of them work less than you'd hope, and a few of them make sleep worse while feeling like they helped. The difference between those categories isn't the price tag. It's whether the thing you're taking actually targets the reason you're awake.

Here's what the evidence says about the common options, and how to test any of them on yourself using the watch you already wear.

TL;DR: Cognitive behavioral therapy for insomnia (CBT-I) is the first-line treatment for chronic insomnia and outperforms medication long term. Melatonin has modest, real evidence for jet lag and shift work, and weak evidence for ordinary insomnia. Antihistamine sleep aids work for a night or two, then tolerance develops within days, and they often leave you groggy. Alcohol is the worst option: it sedates you into lighter, more fragmented sleep. Supplements like magnesium and glycine have small effects at best. If you're waking unrefreshed for months, get evaluated for sleep apnea before buying anything.

YouTube: Related video

The question you should ask before taking anything

"Does it work?" is too vague to answer. A sleep aid can do four different jobs, and almost nothing does all four well:

  1. Get you to sleep faster (sleep latency)
  2. Keep you asleep (wake after sleep onset)
  3. Improve sleep quality (deep sleep, REM, how restorative it feels)
  4. Leave you functional the next day (no grogginess, no driving risk)

Most over-the-counter sleep aids are decent at number one, mediocre at number two, neutral-to-bad at number three, and often bad at number four. Prescription hypnotics can be strong on one and two, with real trade-offs on four and on dependence. Behavioral approaches are slower to work and better on all four.

Keep those four jobs in mind, because "it helped me fall asleep" is a much weaker claim than it sounds.

The options, ranked by evidence

Option Good evidence for Main catch Verdict
CBT-I Chronic insomnia, long-term outcomes Requires effort, takes weeks Strongest option, first-line
Melatonin Jet lag, shift work, delayed sleep phase Weak for ordinary insomnia; timing matters Useful, narrow
Antihistamines (diphenhydramine, doxylamine) Short-term sedation Tolerance in days, next-day impairment Occasional use only
Prescription hypnotics Short-term insomnia Dependence, side effects, limited duration Doctor-guided, short-term
Alcohol Nothing Fragmented sleep, higher overnight heart rate Avoid as a sleep aid
Magnesium, glycine, L-theanine Mild relaxation Small effects, weak trials Low risk, low reward
Valerian, tart cherry, other botanicals Mixed, often negative Inconsistent products Not worth much

Now the detail behind each row.

Antihistamines: real sedation, fast tolerance

Diphenhydramine and doxylamine are the active ingredients in most "PM" products and nighttime cold medicines. They're sedating because they block histamine in the brain, and histamine is part of the system that keeps you awake.

The short-term evidence is genuine. A 1983 double-blind crossover trial in 111 mildly to moderately insomniac patients found 50 mg of diphenhydramine at bedtime improved sleep latency and several other sleep measures compared with placebo. People preferred it, despite more side effects. That's a real effect, not a placebo.

The problem is what happens by day three. A 2002 randomized crossover study gave healthy men 50 mg of diphenhydramine twice daily for four days. On day one, they were measurably sleepier and performed worse on psychomotor tests than placebo. By day four, both the sleepiness and the performance impairment were indistinguishable from placebo. Tolerance developed within three days.

So you get one or two useful nights, then you're taking a drug that mostly isn't doing anything except adding anticholinergic load. And even on night one, the sedation carries over. A separate randomized trial found diphenhydramine produced next-day sleepiness and worse psychomotor performance compared with placebo, which matters if you're driving or operating anything sharp.

There's one more thing worth knowing: first-generation antihistamines are on the Beers list of medications to avoid in older adults, largely because of confusion, falls, and urinary retention risk. If you're over 65, they're a poor choice even occasionally.

Verdict: fine for the night before a flight or the occasional rough night. Not a strategy.

Melatonin: a timing signal, not a knockout pill

Melatonin is the most misunderstood supplement in the cabinet. It's a hormone your pineal gland releases in the evening to signal that it's dark. It is not a sedative. Taking it doesn't switch your brain off; it tells your internal clock what time it is.

That reframing explains both where it works and where it doesn't. The NIH's National Center for Complementary and Integrative Health summarizes the evidence as best for jet lag and shift work, with less certainty for other uses. If your problem is a misaligned body clock, melatonin helps. If your problem is anxiety at 3am, it's not going to do much.

Practical points that actually matter:

  • Timing beats dose. Take it 1–2 hours before your target bedtime, not when you're already frustrated at 2am.
  • Small doses work. Physiologic doses are in the 0.5–3 mg range. More isn't better.
  • Jet lag is its best use case. Take it at the destination's bedtime for the first few nights, and pair it with morning light.
  • Long-term safety is genuinely unknown. NCCIH notes uncertainties about prolonged use, because melatonin is a hormone and could affect other hormonal systems.
  • Supplements aren't tightly regulated. Actual content can vary from the label, so choose third-party tested products.

Verdict: genuinely useful for travel and shift work. Overrated for ordinary difficulty falling asleep.

Prescription sleep medications

This is a doctor conversation, not a blog post, but the framework is worth knowing. The American College of Physicians guideline is clear: CBT-I should be the initial treatment for chronic insomnia, and medication should be considered through shared decision-making when CBT-I alone hasn't worked. Their guidance also notes that medications are generally intended for short-term use, roughly four to five weeks, while the skills learned in CBT-I handle the long term.

That's not because sleep medications are dangerous for everyone. It's because the evidence for long-term benefit is thin, tolerance and dependence are real, and the underlying cause doesn't go away when the prescription ends.

Verdict: useful short-term tool under supervision, especially during acute stress or illness. Not a fix.

Alcohol: the sleep aid that quietly ruins sleep

Alcohol is the most widely used sleep aid in the world and probably the most counterproductive.

It does make you fall asleep faster. It also does three things that make the night worse:

  • It suppresses REM sleep, especially in the first half of the night.
  • It fragments the second half as it metabolizes, which is why you wake at 3am.
  • It raises overnight heart rate and suppresses HRV, which your watch will happily record for you.

If you wear a ring or watch to bed, the pattern is unmistakable: lower HRV, higher resting heart rate, more awake time, a worse sleep score. Our breakdown of what alcohol actually does to sleep and recovery has the full data picture. A nightcap is borrowing sleep from tomorrow at a bad interest rate.

Verdict: not a sleep aid. It's a sedative with a rebound.

Supplements: magnesium, glycine, L-theanine, and friends

These are the least risky options and also the least impressive.

  • Magnesium glycinate is well tolerated and may help people who are genuinely deficient or who report restless legs. The sleep-specific evidence is modest, and studies often use people with insomnia or low magnesium status rather than the general population.
  • Glycine (3 g) has small trials showing improved subjective sleep quality. Effects are subtle.
  • L-theanine produces relaxation without sedation. Useful for winding down, not for passing out.
  • Valerian has inconsistent evidence across trials and wildly variable products.
  • Tart cherry juice has a handful of small studies showing modest melatonin-related effects. Pleasant, not powerful.

The realistic framing: these are low-risk ways to build a wind-down routine. If you sleep better on them, part of the benefit is probably the routine itself, which is fine. The ritual has value.

Verdict: low risk, low reward. Fine as a habit anchor, not a treatment.

What actually outperforms all of it

The intervention with the best long-term evidence isn't a pill. It's CBT-I: a structured program that combines sleep restriction, stimulus control, cognitive work, and sleep hygiene education. It works about as well as medication in the short term and better in the long term, without dependence or next-day impairment.

The core moves are unglamorous:

  • Fix your wake time, seven days a week, before you touch bedtime.
  • Only get into bed when sleepy, and get out if you're awake and frustrated for more than about 20 minutes.
  • Compress your time in bed so your sleep drive is higher when your head hits the pillow.
  • Stop treating the bedroom as an office, a cinema, and a place to argue.

If that sounds like a lot, our guide to CBT-I for insomnia walks through how it works in practice. And if you want to understand the specific problem of lying awake unable to drift off, sleep latency explained covers what's normal and what isn't.

How to test a sleep aid on yourself

This is where having a watch helps. You can run a fairly clean self-experiment in two weeks.

Week 1: change nothing. Record sleep latency, awake time, sleep score, overnight heart rate, and morning HRV. Note how you feel on a 1–10 scale at 10am.

Week 2: add the intervention. Change one thing only. Keep the same bedtime, wake time, caffeine, and training.

Then compare medians, not single nights. Look for three signals:

  1. Does sleep latency drop by more than 10 minutes? That's a real effect.
  2. Does your overnight heart rate or HRV change? This tells you whether the drug is actually altering sleep physiology, not just perception.
  3. Are you more or less functional the next morning? This is the one that decides whether it's worth continuing.

A note on measurement: consumer wearables are decent at detecting sleep versus wake and excellent at tracking heart rate and HRV trends overnight. They're less reliable for deep versus REM staging. So trust your device for "did I sleep" and "how did my body respond," and treat stage breakdowns as approximate. Our piece on wearable sleep tracking accuracy goes deeper.

If a supplement shows no measurable benefit in two weeks, you have your answer.

When to stop self-experimenting

Some sleep problems need a clinician, not a supplement. Book an appointment if you have:

  • Loud snoring, gasping, or witnessed pauses in breathing
  • Daytime sleepiness that's dangerous (falling asleep driving or in meetings)
  • Leg discomfort or an urge to move at night that improves when you walk
  • Insomnia lasting more than three months
  • Sleep problems that started with a new medication
  • Pregnancy, or a condition where supplements could interact with treatment

One useful thing to bring: your watch's overnight respiratory rate and blood oxygen trends. They're not diagnostic, but they can be a helpful prompt for a conversation about sleep apnea. Our summary of wearable hints for sleep apnea explains what those patterns look like.

The honest summary

  • CBT-I beats pills for chronic insomnia, on both short- and long-term outcomes.
  • Melatonin is a timing tool with real evidence for jet lag and shift work, weaker evidence for general insomnia.
  • Antihistamines work for a night or two, then tolerance sets in within days, with next-day impairment.
  • Prescription hypnotics are short-term tools, not solutions.
  • Alcohol makes sleep worse even though it feels like it helps.
  • Supplements are mostly low-risk, low-reward rituals.
  • Measure it. Two weeks of honest data beats five years of guessing.

Sleep is one of the few things where doing less, more consistently, usually beats doing more. Fix your wake time, get light in the morning, keep the room cool, and give behavioral approaches a real trial before reaching for the bottle.


Century AI helps you understand your body with a daily health score, recovery score, and sleep insights — using the watch you already wear.

Tagssleepsleep-aidsmelatonininsomniasleep-quality

Wearable measurements are estimates, not diagnoses. Read Century’s methodology and medical disclaimer.