Sleep medications are not a ladder from weak to strong. A drug that helps someone fall asleep may do little for repeated awakenings. Another may last long enough to help sleep maintenance but also impair driving or balance the next morning.
For chronic insomnia, major guidelines recommend cognitive behavioral therapy for insomnia, or CBT-I, as first-line care. Medication can sometimes be added when benefits are likely to outweigh harms, access to CBT-I is limited, or a person needs short-term support. The choice should be a shared decision, not a search for the most sedating pill 1 2.
This comparison is about medications used for adult insomnia. It does not provide a dose, select a drug for you, or replace an evaluation by a clinician or pharmacist.
Confirm that the problem is insomnia first
Chronic insomnia involves difficulty falling asleep, staying asleep, or waking earlier than intended, along with meaningful daytime distress or impairment. It occurs despite adequate opportunity and circumstances for sleep, typically at least three nights per week for at least three months 1.
Not having enough time to sleep is sleep deprivation, not insomnia. A medicine that produces drowsiness cannot create a safe full sleep opportunity or correct an incompatible work schedule.
Before choosing a medication, the clinical history should look for:
- Loud snoring, breathing pauses, gasping, morning headaches, or marked sleepiness that may point to obstructive sleep apnea
- An urge to move the legs, unpleasant leg sensations, or nighttime limb movements
- A sleep schedule that is delayed, advanced, irregular, or misaligned with shift work
- Pain, reflux, urinary symptoms, hot flashes, pregnancy-related changes, or another medical contributor
- Depression, anxiety, post-traumatic stress, mania, or another mental health condition
- Caffeine, nicotine, alcohol, cannabis, stimulants, opioids, and other substances
- Prescription drugs, over-the-counter products, and supplements that may disrupt sleep or add sedation
- Whether there is enough time and a suitable environment for sleep
Insomnia and another condition can coexist. Treating apnea, restless legs syndrome, pain, depression, or a circadian disorder does not make the insomnia complaint unimportant. It changes what treatment is likely to help and what may be unsafe.
CBT-I comes before routine medication
CBT-I is more than general sleep-hygiene advice. It combines a structured sleep schedule, stimulus control, cognitive strategies, and other techniques that address processes maintaining insomnia. The 2025 VA/DoD guideline recommends CBT-I and suggests it over medication as first-line treatment. The American College of Physicians also recommends CBT-I first, followed by shared decision-making if medication is considered 1 2.
Medication and CBT-I do different jobs. A sedative may change sleep on nights it is taken. CBT-I teaches skills intended to remain useful after treatment ends. Some people use both, but taking a pill should not make access to CBT-I or evaluation of another sleep disorder optional.
Sleep onset and sleep maintenance are different targets
Sleep-onset insomnia means difficulty falling asleep at the intended time. Sleep-maintenance insomnia means repeated or prolonged awakenings, difficulty returning to sleep, or waking too early. Many people have both.
FDA approval for one target does not prove benefit for the other. Formulation matters too. Immediate-release and extended-release versions of the same ingredient can have different indications and next-day effects.
| Medication option | Adult insomnia target in U.S. labeling or guidance | Main decision points |
|---|---|---|
| Dual orexin receptor antagonists: daridorexant, lemborexant, suvorexant | Sleep onset and/or sleep maintenance | Schedule IV drugs; next-day impairment, falls, unusual sleep experiences, interactions, narcolepsy, and respiratory status matter |
| Z-drugs: eszopiclone, zaleplon, zolpidem | Eszopiclone: onset and maintenance; zaleplon: onset; zolpidem: target depends on formulation | FDA boxed warning for complex sleep behaviors; next-day impairment, falls, dependence, and formulation-specific directions matter |
| Benzodiazepines used for insomnia | Target and labeled duration vary by drug | FDA boxed warning for abuse, misuse, addiction, physical dependence, and withdrawal; current VA/DoD guidance suggests against routine use for chronic insomnia |
| Low-dose doxepin | Sleep maintenance | Not a controlled substance; sedation, interactions, urinary or eye conditions, mood, and respiratory status matter |
| Ramelteon | Sleep onset | Not a controlled substance; evidence is mixed, and liver function and drug interactions matter |
The table summarizes treatment targets, not comparative safety or effectiveness. A large network meta-analysis found meaningful differences among medications, but direct long-term comparisons were sparse and confidence varied by drug and outcome 3.
Many trials report average changes in time to fall asleep or time awake after sleep onset measured in minutes. Daytime function does not always improve alongside a sleep measure, and a group average cannot predict one person's response 1 3.
FDA-approved prescription categories
Dual orexin receptor antagonists
Daridorexant, lemborexant, and suvorexant reduce signaling through orexin receptors, part of the brain's wake-promoting system. Their current U.S. labels cover insomnia with difficulty falling asleep and/or staying asleep 4 5 6.
These drugs are Schedule IV controlled substances. Their labels warn about daytime sleepiness or impaired alertness, falls, complex sleep behaviors, sleep paralysis, vivid perceptions around sleep transitions, and symptoms resembling brief muscle weakness. All three are contraindicated in narcolepsy.
The drugs should not be treated as interchangeable. Their metabolism, interaction instructions, liver-disease limits, studied populations, and respiratory data differ. For example, a breathing study performed with one orexin drug cannot establish safety for another, and a study in mild apnea cannot answer every question about severe untreated apnea or advanced lung disease. The applicable product label and the person's full medication list should guide the discussion.
The Z-drug boxed warning does not automatically apply to orexin drugs. Orexin-drug labels still contain their own warnings about complex sleep behaviors, but not the same FDA boxed warning.
Nonbenzodiazepine receptor agonists, or Z-drugs
Eszopiclone, zaleplon, and zolpidem act at the benzodiazepine site of the GABA-A receptor, although they are not chemically benzodiazepines. Their targets are not identical:
- Eszopiclone has evidence and labeling for reducing sleep latency and improving sleep maintenance 7.
- Zaleplon is a short-term sleep-onset treatment. Its label says it has not been shown to increase total sleep time or reduce awakenings 8.
- Immediate-release zolpidem is labeled for short-term sleep-initiation difficulty, while extended-release zolpidem covers sleep onset and maintenance. These formulations have different instructions and cannot be substituted based only on the ingredient name 9 10.
The FDA requires a boxed warning for all three Z-drugs because complex sleep behaviors, including sleepwalking, sleep-driving, cooking, or other activity while not fully awake, have caused serious injury and death. Events have occurred after a first dose and at recommended doses. A previous complex sleep behavior after taking one of these medicines is a contraindication to this group of Z-drugs 11.
Z-drugs can also cause next-day impairment, memory or coordination problems, falls, and dependence. A person may not accurately judge their own driving impairment. Sufficient time for sleep, the exact formulation, other sedatives, liver function, age, and morning responsibilities all affect risk.
Benzodiazepines used for insomnia
Some benzodiazepines, including temazepam and triazolam, have insomnia indications. Their treatment targets and labeled durations differ, but the 2025 VA/DoD guideline suggests against benzodiazepines for chronic insomnia because the balance of evidence and harms is unfavorable 1.
All benzodiazepines carry an FDA boxed warning about abuse, misuse, addiction, physical dependence, and withdrawal. Dependence can occur even when a medicine is taken as prescribed. Combining a benzodiazepine with an opioid, alcohol, or another central nervous system depressant can cause profound sedation, respiratory depression, coma, or death 12.
Abrupt discontinuation or a rapid reduction after regular use can cause dangerous withdrawal, including seizures. There is no single taper schedule that suits everyone. A clinician should build an individualized plan based on the drug, duration, pattern of use, health conditions, and symptoms.
Low-dose doxepin
Doxepin is a tricyclic antidepressant, but the low-dose insomnia product primarily blocks histamine signaling and is FDA-approved for sleep-maintenance insomnia. It is not a controlled substance 13.
The insomnia formulation should not be treated as equivalent to higher antidepressant doses. Higher-dose doxepin has a different adverse-effect burden, and evidence from one dose range should not be transferred to the other. Even at an insomnia dose, next-day sedation, additive effects with alcohol or sedating drugs, mood changes, severe sleep apnea, urinary retention, narrow-angle glaucoma, and drug interactions can affect suitability.
Ramelteon
Ramelteon acts at melatonin receptors and is FDA-approved for difficulty with sleep onset. It is not a controlled substance 14.
The 2025 VA/DoD panel found insufficient evidence to recommend for or against ramelteon for chronic insomnia. That is different from saying it never works. Trials show some improvement in sleep-onset measures, but results have not been consistent across subjective sleep, daytime function, and other outcomes 1.
Ramelteon has product-specific interaction and liver-disease restrictions. It may still cause drowsiness or affect driving. Prescription ramelteon should not be confused with over-the-counter melatonin, which has different ingredients, regulation, and evidence.
Melatonin and over-the-counter antihistamines
Melatonin supplements
Melatonin is a hormone signal involved in circadian timing. In the United States, supplements are not FDA-approved medicines for chronic insomnia, and product content may differ from the label. Current guidance does not find strong enough evidence to recommend melatonin as a routine treatment for chronic insomnia 1 15.
This conclusion does not apply to every circadian problem. Strategically timed melatonin may have a role in selected circadian rhythm sleep-wake disorders or jet lag. That is a timing treatment for a different problem, not evidence that more melatonin is a stronger insomnia treatment.
Melatonin can cause sleepiness, headache, dizziness, or nausea and can interact with medicines. Long-term safety data are limited. Pregnancy, breastfeeding, epilepsy, blood thinners, and use in a child are reasons to involve a qualified clinician rather than self-treat 15.
Diphenhydramine and doxylamine
First-generation antihistamines can produce drowsiness, which is why diphenhydramine and doxylamine appear in many nighttime products. Sedation is not the same as treating chronic insomnia. Trials of diphenhydramine have shown mixed or small benefits, and tolerance to the sedating effect may develop 1.
Possible effects include next-day drowsiness, dry mouth, constipation, blurred vision, urinary retention, confusion, and impaired coordination. The same ingredient may appear in an allergy, cold, pain, or “PM” product, creating a risk of unintended duplication.
These anticholinergic effects are especially concerning in older adults. The AGS Beers Criteria advises avoiding first-generation antihistamines and also flags benzodiazepines and Z-drugs because of risks such as confusion, delirium, falls, fractures, and motor vehicle crashes 16.
Common off-label medications
Off-label means a medicine is being used for a purpose not approved in its FDA labeling. It does not automatically mean the use is wrong, but sedation alone is not proof that a drug effectively treats chronic insomnia.
Trazodone and other sedating antidepressants
Trazodone is approved for depression, not insomnia. The 2025 VA/DoD guideline suggests against using it for chronic insomnia. The reviewed trials were short and did not show consistent improvements across sleep latency, wake time after sleep onset, sleep efficiency, and daytime function. Morning grogginess, dry mouth, low blood pressure, falls, priapism, interactions, and the antidepressant boxed warning about suicidal thinking in younger people are part of the decision 1.
Mirtazapine, amitriptyline, and other sedating antidepressants also lack strong evidence as routine treatments for primary chronic insomnia. They may be prescribed when depression, neuropathic pain, migraine, or another approved or evidence-supported target is present. That is a different decision from using them only because they cause drowsiness. Weight change, metabolic effects, anticholinergic effects, blood-pressure changes, and daytime sedation differ by drug.
Antipsychotics
Quetiapine and other antipsychotics are not approved as insomnia medicines. Evidence for primary insomnia is very limited, while metabolic, cardiovascular, movement, blood-pressure, cognitive, and fall risks can be substantial. Antipsychotics also carry a boxed warning about increased mortality in older adults with dementia-related psychosis. Current VA/DoD guidance suggests against using them for chronic insomnia 1.
A psychiatric indication can justify an antipsychotic in an individualized treatment plan. It does not turn the drug into a generally supported insomnia treatment.
Gabapentinoids
Gabapentin and pregabalin are not FDA-approved treatments for insomnia. They may be used for specific pain, seizure, or restless legs indications, depending on the product and clinical situation. Any improvement in sleep while treating that condition should not be assumed to prove benefit for primary insomnia.
The FDA warns that gabapentinoids can cause serious breathing problems in people with respiratory risk factors, including those using opioids or other central nervous system depressants, people with impaired lung function, and older adults 17.
Safety questions that can change the choice
Next-day alertness, driving, falls, and cognition
Any sedating medicine can affect reaction time, judgment, balance, or memory after waking. Risk depends on the drug, formulation, timing, metabolism, sleep opportunity, and other substances. Feeling awake is not proof that driving performance is normal.
Before treatment, discuss early shifts, caregiving, commercial driving, operating machinery, nighttime bathroom trips, fall history, and cognitive impairment. Do not drive or perform hazardous work when sleepy or impaired. New confusion, repeated falls, or dangerous next-day sedation warrants prompt review.
Alcohol, opioids, and other sedatives
Alcohol is not an insomnia treatment and should not be combined with a sleep medicine. Opioids, benzodiazepines, gabapentinoids, sedating antihistamines, muscle relaxants, some antidepressants, antipsychotics, and other sleep products can have additive effects.
Bring one complete list of medicines, supplements, cannabis products, and alcohol use to the prescriber and pharmacist. Checking ingredients matters because combination products can hide diphenhydramine, doxylamine, acetaminophen, or another duplicated drug.
Extreme sleepiness, slowed or difficult breathing, blue or gray lips, inability to wake, a seizure, or a suspected overdose needs emergency help.
Sleep apnea and respiratory disease
A sedative does not treat airway collapse. Untreated sleep apnea, COPD, hypoventilation, neuromuscular disease, opioids, and multiple sedatives can change respiratory risk.
This is not a rule that every insomnia medication is forbidden with apnea. Drug labels and respiratory studies differ. The right approach is to identify and treat the breathing disorder, then use product-specific evidence rather than assuming a class is safe or unsafe.
Older adults
Older adults may clear medicines more slowly and can be more sensitive to sedation, confusion, low blood pressure, and balance effects. The Beers Criteria generally advises avoiding benzodiazepines, Z-drugs, and first-generation antihistamines in this population. It distinguishes low-dose doxepin used for insomnia from higher antidepressant dosing 16.
Beers Criteria are a safety screen, not an automatic command to stop a medicine. Current health, treatment response, alternatives, and withdrawal risk still need individual review.
Pregnancy, breastfeeding, children, and adolescents
Pregnancy and the postpartum period can change sleep, breathing, restless legs symptoms, reflux, mood, and medication handling. Pregnancy and lactation data differ by product and are often limited. Review reproductive status before prescribing rather than declaring one class universally safest 1.
This adult comparison should not be used to choose a sleep medicine for a child or teenager. Pediatric insomnia has different causes and evidence, and prescription labels have age-specific limits. Even melatonin should be discussed with a pediatric health professional because long-term data and product consistency are limited 15.
Storage, sharing, and overdose
Keep prescription drugs, antihistamines, and supplements in their original containers and out of sight and reach of children. Controlled medicines should be secured, never shared, and disposed of through an appropriate medication take-back option when no longer needed.
If depression, suicidal thinking, substance misuse, memory problems, or accidental double-dosing is a concern, the prescriber can consider quantity, packaging, caregiver support, and follow-up. “Over the counter” and “natural” do not mean harmless in an overdose.
Duration and follow-up are product-specific
There is no responsible universal rule that every insomnia medicine must stop on the same day. Some labels describe short-term use, while trials supporting other products lasted several months. Long-term comparative evidence remains much thinner than short-term evidence 3.
A medication plan should define:
- The target, such as time to fall asleep, prolonged awakenings, or daytime function
- How benefit and next-day impairment will be tracked
- When the first review will occur
- What would prompt earlier review, such as a fall, complex sleep behavior, mood change, or breathing concern
- Whether CBT-I and treatment of contributing conditions are progressing
- How the medicine will be reduced or stopped if risks outweigh benefits
Lack of improvement, worsening insomnia, escalating use, or continued impairment should prompt reassessment rather than automatic dose increases or medication stacking. Do not improvise a taper or abruptly stop a regularly used sedative, particularly a benzodiazepine. Withdrawal and rebound risk depend on the specific drug and use pattern 12.
Questions to bring to the medication discussion
A useful conversation with a clinician or pharmacist can start with:
- Is this chronic insomnia, short-term insomnia, insufficient sleep, or another sleep disorder?
- Is my main target sleep onset, sleep maintenance, or both?
- What benefit did trials show for this exact drug and formulation, and did they measure daytime function?
- What are the next-day, fall, cognitive, breathing, and complex-behavior risks for me?
- Does it interact with alcohol, opioids, my other medicines, or supplements?
- Is it a controlled substance, and what are the dependence and withdrawal considerations?
- When will we review benefit and harm, and what is the plan if it does not help?
- How can I access CBT-I or a suitable alternative format?
The bottom line
The best medication decision is not the one that produces the most sedation. It is the one that matches a confirmed treatment target, has evidence relevant to that target, and has acceptable risks for the person's health, other medicines, and next-day responsibilities.
CBT-I remains first-line care for chronic insomnia. If medication is used, its benefit should be checked against daytime function and safety, with product-specific follow-up and a clinician-guided discontinuation plan. Trial averages can inform that decision, but they cannot predict which individual will benefit or make one medicine universally safest.
