You've probably tried a lot already: cutting the caffeine, a bedtime routine, a podcast in the dark. So reaching for a pill can feel like failure and relief at once, and knowing the pack is there in the drawer may be half the comfort. If you've taken one for a long time, you may have a worry behind all the other questions. What is it doing to you, and could you sleep without it? For tonight, though, it comes down to some sleep, and a clear head when the alarm goes tomorrow.

How much does a sleeping pill add to a night?

What most people hope for from a pill is simple enough: to fall asleep sooner, and to sleep longer once they do. The prescription pills you're most likely to be holding, zolpidem or an older benzodiazepine such as temazepam, work through the brain's own braking system35 to slow the brain and bring on drowsiness.

Researchers once gathered the trials that the makers of zolpidem and its relatives had sent to the FDA: thirteen of them, with more than 4,000 adults with insomnia sleeping in a lab. On the drug, people fell asleep about forty-two minutes sooner5 than they had before. People on a dummy pill fell asleep sooner too. Set side by side, about twenty-two of those minutes came from the drug itself5, and about twenty came on the dummy pill as well.

What, then, does the pill itself add to your night? The American Academy of Sleep Medicine pooled the zolpidem trials for its guideline. Taken together, they showed zolpidem cut about twelve minutes from falling asleep and added about half an hour of sleep2. The academy rates the evidence behind those numbers as very low in quality, so they're best read as a rough guide. Why twelve minutes here and twenty-two in the FDA trials? The two figures come from different sets of trials. For the older benzodiazepines, the trials are thinner still2. Newer pills that act on the brain's orexin system, such as daridorexant, do modestly better than a dummy pill6, with no evidence so far of tolerance or rebound26 when they're stopped. Those trials ran only one to three months, though, which is too short to know for sure.

So on a hard night a pill can give you something real: somewhere between twelve and twenty-two minutes sooner asleep from the drug itself, and about half an hour more sleep. A good part of the ease you feel on it comes from you.

When thou liest down, thou shalt not be afraid: yea, thou shalt lie down, and thy sleep shall be sweet.
Proverbs, by The Bible, King James Version 3:24

What about the PM pills in the drugstore?

In many PM boxes, among them Advil PM, Aleve PM and Excedrin PM, the sleep ingredient is diphenhydramine, the same antihistamine sold for allergies as Benadryl7. That makes it easy to take twice in one evening without knowing: once in a painkiller, and again in an allergy tablet. Some PM boxes use a different ingredient, so you could turn yours over and check the label. What follows is about diphenhydramine alone.

On a hard night, a PM pill is likely to add only minutes of sleep2, too few for the sleep medicine guideline to count as a real difference. Your body also seems to get used to it within days. In a small trial, fifteen healthy men took it twice a day, and the sleepiness it caused had faded by the fourth day4. That was sleepiness by day; the trial didn't look at anyone's nights. If you have glaucoma, lung disease or trouble passing urine because of an enlarged prostate, you'd want a doctor to hear about it before you take it7.

Extra sleep over a dummy pill, and how far each guess might be off

Each dot is the best guess, in the trials the American Academy of Sleep Medicine pooled, of how much more adults with insomnia slept on the pill than on a dummy pill; each line is the range the true gain could plausibly fall in. Zolpidem’s whole range is above zero, measured in a sleep lab or by people’s own reckoning2. The lines for triazolam, from a single trial of 32 people2, and for diphenhydramine, from two trials2, reach past no difference: they can’t rule out adding nothing at all.

From the studies’ own measurements

See this as a table
Difference in minutes, Less sleep on the pill or More sleep on the pill; zero is No difference
MeasureBest estimateLikely rangeNote
Zolpidem, measured in a sleep lab 29 minutes more 11 more to 47 more Two lab studies
Zolpidem, by people’s own reckoning 30 minutes more 15 more to 45 more Pooled trials
Triazolam, by people’s own reckoning 25 minutes more 9 less to 60 more One trial
Diphenhydramine (PM pills), own reckoning 18 minutes more 4 less to 40 more Two trials

Is it all right to take one on a hard night?

Taken now and then, a pill is a visitor your body barely adjusts to, because dependence builds with steady use3. So on an occasional night, what matters most is how and when you take it. Zolpidem's label says to take it right at bedtime, with seven to eight hours still ahead of you in bed1. If you get up sooner, you can be drowsy, unsteady and forgetful, and it can impair your driving the next morning even when you feel wide awake1. The guideline found this morning-after sleepiness was the main exception to zolpidem's otherwise mild side effects2, most of all with fewer than eight hours before waking. That's why a pill taken at 3am with the alarm set for 7 is the one most likely to follow you into the day.

The serious danger lies in combinations, when a pill meets other things that slow the brain and breathing. Benzodiazepines taken with opioids, alcohol or other sedatives have caused breathing so slow that people have died3. Alcohol adds to the effects of zolpidem and diphenhydramine17 as well. So please don't take a sleeping pill on a night you've been drinking, or alongside another sleep aid. And if you wake at 2am with the pill already in you, please leave the pack where it is rather than reach for a second one1.

A few people on zolpidem have got up and driven, cooked and eaten, made phone calls or sleepwalked without being fully awake1, and remembered none of it the next day. Trials are too small to say how often this happens, so what's known comes mostly from clinical experience2. If you find it has happened to you, please call your doctor straight away. The label says to stop the pill. If you've been taking it every night, your doctor can tell you how to stop safely, since stopping all at once can bring withdrawal.

Is it different if I’m older?

If you're older, or helping a parent decide, the pills work differently in an older body. Here the evidence comes from trials in people over 60, who took a pill for at least five nights. Sleeping pills gave them about twenty-five more minutes of sleep a night8. Memory loss, confusion and daytime tiredness, meanwhile, came far more often than on a dummy pill. For every thirteen people treated, one slept better, while for every six, one had a side effect8.

That's why the American Geriatrics Society advises older adults to avoid benzodiazepines for sleep and keep z-drugs to under ninety days2, because of falls, confusion and car crashes. Zolpidem can leave someone unsteady in the night and the next day, so getting up to the bathroom at night and the first hours of the morning both need extra care1. PM pills matter here too. A box on an open shelf isn't a sign that it suits an older body: diphenhydramine is generally not advised for people 65 and over7 except for a serious allergic reaction.

What happens if I take one every night?

If you've taken a pill most nights for months or years, nothing here asks you to stop tonight. Long-term use can be reasonable. The guideline notes that newer z-type drugs can stay safe and effective over the long term2 for some people: those for whom CBT for insomnia hasn't worked or isn't available, who keep benefiting and who see their doctor regularly.

Taken every night, though, a pill changes how your body responds to it. Zolpidem may not help as well after two weeks or more1 as it did at first. With a benzodiazepine, a few days to a few weeks of steady use, even exactly as prescribed, is enough for the body to adapt to the drug3. Then stopping suddenly brings withdrawal. This is physical dependence, and it's a different thing from addiction: your body has rearranged itself around a pill that arrives every night. How often either happens with sleeping pills isn't well known23. The risk of dependence is greatest with benzodiazepines taken for a long time, at rising doses, with no one keeping an eye on them.

It’s often said

I can’t sleep without it.

Why it can feel that way

If you take a pill every night, you may remember the night you tried without it: hour after hour wide awake, worse than the insomnia you started with, until it felt like proof that only the pill was getting you to sleep.

What’s going on underneath

A sudden stop can bring rebound insomnia, a spell of worse sleep once the pill is gone. After zolpidem it's mostly limited to the first night2, and some studies found none at all after up to six months of use. The label says it usually passes on its own within one or two nights1. Coming off a benzodiazepine after years is a different matter, and for some people withdrawal brings anxiety, poor sleep and low mood that linger beyond four to six weeks, for weeks more or as long as a year3. Since zolpidem's rebound mostly passes within a night or two, a hard night after a sudden stop says more about the stop than about how you'd sleep once you'd got used to going without the pill.

If the pill you take every night is a PM one, the evidence is a separate and thinner story. The sleep medicine guideline suggests not using diphenhydramine for long-running insomnia, because in trials it added only minutes of sleep2. In that small trial, the daytime sleepiness it caused in healthy men had faded by the fourth day4, a hint rather than a measure of anyone's nights. And heavy use over many years9 is where the early dementia link you'll read about below comes from.

If you're 65 or over, diphenhydramine is generally not advised7, and your doctor would want to know you take it. Whatever your age, a pharmacist is someone you could talk it over with, or your doctor if you have one.

How do people come off sleeping pills?

Please don't stop a nightly benzodiazepine or z-drug all at once. Stopping suddenly, or cutting down too fast, can bring withdrawal: with zolpidem, shakiness, nausea, sweating, panic, bad nights and, rarely, seizures1, and with benzodiazepines, seizures that can be life-threatening3. The way off is a gradual taper you plan with your doctor3, stepping the dose down over time. No single schedule suits everyone. If withdrawal shows up along the way, the two of you can pause the taper or step it back, then carry on more slowly.

In Québec, researchers worked with seventy-six older adults who had taken a benzodiazepine for sleep for an average of nineteen years. Over ten weeks, some tapered under a doctor's supervision, some had CBT for insomnia alongside the same taper, and some had CBT for insomnia alone. By the end, eighty-five percent of those who had both were off the drug10, against forty-eight percent with the taper alone and fifty-four percent with CBT alone. A year later, they had kept those reductions. They had no significant withdrawal symptoms, and their sleep went on improving for months.

So if you're looking for a way off, what helps most? When researchers pooled eight trials like this one, people who added CBT for insomnia to their taper were more likely to be off their benzodiazepine sleeping pills11 in the first few months. By a year, the difference was no longer clear. A 2025 review found low-certainty evidence that simpler help may count too: explaining the pills and the taper, a look through someone's medicines, or a pharmacist's program may help more people stop12. That same review didn't find compelling evidence for CBT on its own12 as a way off. The two reviews asked different questions: one compared CBT added to a taper with the taper alone, and the other looked at CBT by itself. Neither settles which kind of extra help works best. A reasonable way to go, then, is a slow taper planned with your doctor, with CBT for insomnia alongside it if you can get it.

Guidelines put CBT for insomnia first213 for long-running insomnia, because its gains last after the sessions end, while a pill's tend to fade once you stop. If you'd like a next step, you could read what it would ask of you, whether or not you're ready to come off.

Do sleeping pills cause dementia?

It's a frightening thing to read, and researchers don't agree yet. When studies that followed people over time were pooled, those who took benzodiazepines were about 1.4 times as likely to be diagnosed with dementia14. But which came first, the pill or the dementia? Insomnia and anxiety can be early signs of dementia15, so some people may reach for a pill in its first years, before anyone knows it's there.

A team in Seattle tried to pull those apart. They followed 3,434 people aged 65 and over, and they left out each person's most recent year of pill use, so that early signs of dementia would be less likely to muddy things. If the pills caused dementia, you'd expect the risk to climb with heavier use. It didn't. The risk was slightly higher at the lowest use and not at the highest15, and the researchers said that pattern doesn't support cause and effect.

The PM pills raise a different question. The same Seattle researchers looked at anticholinergic medicines, a group that includes diphenhydramine. Dementia risk was about 1.5 times higher9 in people who'd taken the equivalent of more than three years of daily use over a decade, with no clear rise at lower amounts. That study pooled every kind of anticholinergic medicine, sleep aids among many. It watched people over time without testing a drug, so it's an early finding.

So where does that leave you? There is a link, and researchers are still arguing over it. So far, though, the evidence doesn't show that the pills cause dementia, and the most careful study didn't find the pattern you'd expect if they did. Studies of whole populations have also tied sleeping pills to infection, depression and earlier death in older people, but the guideline calls that evidence scant and inconsistent2. If the worry stays with you, your doctor is a good person to talk it through with. And if you decide to come off, the slow taper above is the way to do it, since stopping all at once can bring withdrawal.

What the research says

  • Z-drugs help people fall asleep about twenty-two minutes sooner than a placebo in trials, out of about forty-two minutes in all; the other twenty or so are the placebo response. Good evidence
  • Diphenhydramine, the sleep ingredient in many PM pills, adds only a few minutes to falling asleep and less than twenty minutes of sleep in trials. Good evidence
  • In one small trial of healthy men, the daytime sleepiness diphenhydramine caused faded within four days, which suggests tolerance comes quickly. Early evidence
  • Combining benzodiazepines with opioids, alcohol or other sedatives can slow breathing dangerously, and almost all US overdose deaths involving benzodiazepines involve other drugs too. Well established
  • Steady use of a benzodiazepine can lead to physical dependence within days to weeks, and stopping suddenly can bring withdrawal, including seizures. Well established
  • Rebound insomnia after stopping zolpidem, when it comes, mostly lasts a night or two. Good evidence
  • A gradual taper with CBT for insomnia helps more people come off benzodiazepines in the first months than a taper alone. Good evidence
  • In people over 60, sleeping pills bring side effects such as confusion about twice as often as they bring better sleep. Good evidence
  • It isn't clear yet whether benzodiazepines raise the risk of dementia, and findings differ: pooled observational studies find a link, while in a careful cohort the risk didn't rise with heavier use. Still debated
  • Heavy, long use of anticholinergic medicines, a group that includes diphenhydramine, was linked with more dementia in one large study of older adults. Early evidence

The sleeping-pill trials pooled here, the z-drug trials submitted to the FDA and the guideline's trials of zolpidem, were paid for by their makers and mostly lasted only weeks, and trials like these are too small to show rare harms2 such as sleep-driving. How often dependence and addiction happen is still not well known23.

On an odd hard night, with a full night in bed ahead and no alcohol, a pill asks very little of you. Every night, it asks more. The way back takes time, but people do find it, even after many years on the pills, as the older adults in Québec10 did.