How to Stop Having Nightmares: What Actually Works

The intervention with the strongest evidence for reducing nightmares is a cognitive-behavioral technique called imagery rehearsal therapy, or IRT, which involves rewriting the script of a recurring nightmare while you’re awake and then mentally rehearsing the new version. It sounds almost too simple, but multiple trials and a meta-analysis show large, lasting reductions in nightmare frequency. Beyond IRT, though, the full picture of what works includes addressing medications that may be triggering your nightmares, managing the stress and sleep problems that feed them, and knowing which newer approaches are gaining ground. Nightmares are more treatable than most people realize, and the most effective options don’t require a prescription.

When Nightmares Cross the Line From Normal to Problem

Disturbing dreams are common. In community surveys, roughly one in twenty adults reports nightmares at least once a week, and the rate runs higher in younger people: a large epidemiological study in Hong Kong found that about 16% of young adults had frequent nightmares.¹² A key distinction researchers draw is between a nightmare, which wakes you up, and a “bad dream,” which is just as unpleasant but doesn’t jolt you out of sleep.³ That waking criterion matters clinically because nightmares that disrupt your sleep repeatedly are the ones that cascade into daytime fatigue, anxiety about going to bed, and worsening mental health.

Frequent nightmares are strongly linked to depression, anxiety, and PTSD. In the Hong Kong study, young adults with frequent nightmares were roughly twice as likely to have moderate-to-severe depressive or anxiety symptoms, and nearly four times as likely to meet criteria for probable PTSD, even after adjusting for other risk factors.² People who had nightmares before experiencing a traumatic event also tended to develop more severe PTSD afterward, suggesting that nightmares aren’t just a symptom of psychological distress but can worsen it.⁴ This is why researchers increasingly treat nightmares as a standalone problem worth addressing directly, rather than waiting for the underlying condition to improve on its own.

Imagery Rehearsal Therapy, the Best-Supported Treatment

IRT is the technique sleep clinicians reach for first. The idea is straightforward: you pick a recurring nightmare, write it down, then change the storyline to something less distressing. You don’t have to make it cheerful; just different enough that it no longer follows the same threatening script. Then you spend ten to twenty minutes a day mentally rehearsing the rewritten dream, ideally before bed. Over days and weeks, the new version tends to crowd out the old one.

The evidence behind IRT is genuinely strong. A meta-analysis covering post-trauma nightmares found large effects on nightmare frequency, sleep quality, and PTSD symptoms from pre- to post-treatment, with improvements sustained through six to twelve months of follow-up.⁵ One of the earliest controlled trials found that treated participants dropped by an average of about two nightmare-nights per week and about four total nightmares per week, with self-rated sleep quality also improving significantly.⁶ The technique works for nightmares tied to trauma and for those that seem to come out of nowhere, which researchers call “idiopathic” nightmares.

A related approach called Exposure, Relaxation, and Rescripting Therapy (ERRT) adds components: you expose yourself to the nightmare content more deliberately (by writing about and sitting with the distressing images), practice relaxation skills, and then rescript the dream. A randomized trial found that ERRT improved nightmare frequency and severity, depression, sleep quality and quantity, and overall quality of life, with gains holding at six months.⁷ The researchers also tracked physiological markers and found that people whose heart rate and skin conductance to nightmare imagery decreased during treatment were the ones whose sleep and PTSD symptoms improved the most. In other words, the body’s alarm response was genuinely quieting down, not just being intellectually overridden.

Combining Nightmare Therapy With Insomnia Treatment

Nightmares and insomnia frequently coexist, and treating one without the other can feel like mopping the floor while the faucet is still running. A group therapy trial that combined cognitive-behavioral therapy for insomnia (CBT-I) with IRT found that participants experienced significant reductions in insomnia severity, nightmare distress and frequency, and PTSD symptoms all at once.⁸ Sleep-diary data showed shorter time to fall asleep, longer total sleep, less time lying awake after initially falling asleep, and higher overall sleep efficiency. Crucially, the degree of insomnia improvement was positively correlated with the degree of nightmare distress reduction, suggesting these two problems reinforce each other and benefit from being tackled together.

If you have both nightmares and trouble falling or staying asleep, asking a clinician specifically about combined CBT-I/IRT is worth doing. Standard CBT-I addresses habits and thought patterns around sleep (stimulus control, sleep restriction, cognitive restructuring of catastrophic sleep beliefs), and layering IRT on top gives you a tool for the nightmare content itself.

Medications That Can Help, and the Prazosin Controversy

The most talked-about drug for trauma-related nightmares is prazosin, a blood-pressure medication that blocks norepinephrine receptors. The theory is that by dampening the body’s fight-or-flight signaling during sleep, prazosin can reduce the intensity and frequency of nightmares. Multiple reviews have found it effective for PTSD-related nightmares.⁹ A systematic review of randomized controlled trials reported that five out of seven concluded prazosin reduced nightmares and improved sleep quality.¹⁰

But the picture got complicated. A large, well-designed trial of prazosin in military veterans with chronic PTSD, published in the New England Journal of Medicine, found no significant difference between prazosin and placebo on distressing dreams or sleep quality at ten weeks or twenty-six weeks.¹¹ This trial was bigger and methodologically stronger than the earlier positive studies, and it rattled the field. Some researchers have speculated the differences may come down to the severity and chronicity of the PTSD, the specific population studied, or the degree of placebo response. The bottom line is that prazosin helps some people with trauma-related nightmares, but it isn’t the slam dunk earlier literature suggested, and it’s not a first-line option for nightmares unrelated to PTSD. It also causes dizziness and low blood pressure in some users, so any trial should be medically supervised.

Medications That Cause Nightmares

Before adding a medication to treat nightmares, it’s worth asking whether one of your current medications might be causing them. Several common drug classes are linked to increased nightmare frequency. Reviews have identified beta-blockers (used for blood pressure and heart conditions), sedative-hypnotics (including some sleep aids, which is ironic), amphetamines, and dopamine agonists (used for Parkinson’s disease and restless legs syndrome) as the categories most frequently associated with nightmares.¹² More broadly, drugs that affect norepinephrine, serotonin, and dopamine have clear associations with nightmare reports, and agents that affect the immune response to infection (some antimalarials and antivirals) can also be culprits.¹³

Withdrawal from certain psychotropic drugs can also trigger nightmares, particularly abrupt discontinuation of antidepressants or benzodiazepines.¹⁴ If your nightmares started or worsened around the time you began or stopped a medication, that’s a conversation worth having with your prescriber. Sometimes a dose adjustment or switch to a different drug in the same class is enough to resolve the problem without compromising the original treatment.

Food, Stress, and Other Everyday Triggers

People have blamed food for bad dreams since at least the Victorian era, and the idea isn’t entirely folklore. In one study, roughly 18% of participants believed certain foods affected their dreaming, and dairy products were the most commonly blamed category. People who perceived food-dependent dreaming also tended to have more disturbing dreams overall, poorer sleep, higher coffee intake, and patterns consistent with emotional eating.¹⁵ The mechanism is debatable. Eating heavily before bed can disrupt sleep architecture, and disrupted sleep means more awakenings during REM, which means more remembered dreams, including the unpleasant ones. Spicy food and alcohol similarly fragment sleep. Whether the food itself changes dream content or simply increases the odds you’ll wake up mid-nightmare is still unclear.

Stress is a more direct trigger. The theoretical framework that many nightmare researchers use proposes that nightmares occur when the brain’s emotional processing system is overloaded. The hippocampus builds the dream’s setting and storyline while the amygdala and prefrontal cortex engage in a kind of tug-of-war over emotional intensity. When waking stress or trauma has piled up enough unresolved emotional material, the normal fear-processing function of dreaming essentially fails, and the result is a nightmare.¹⁶ This model explains why both trauma survivors and chronically stressed people are vulnerable, and why stress reduction techniques (from regular exercise to progressive muscle relaxation before bed) often reduce nightmare frequency even without directly targeting dream content.

Emerging Approaches Worth Watching

One of the more creative recent developments pairs IRT with targeted memory reactivation, or TMR. In a trial with 36 nightmare disorder patients, participants were divided into two groups. Both did IRT, but one group also heard a specific sound while they were creating the positive version of their nightmare. During the two-week home practice period, all participants wore a wireless headband that detected REM sleep. When the headband detected REM, it played the sound. The idea was that the sound would reactivate the rewritten dream memory during the exact sleep stage when nightmares occur. After two weeks, the TMR group had fewer nightmares and more positive dream emotions than the control group, and the reduction in nightmares was still present three months later.¹⁷ This is still a small trial, but the concept of using sound cues during sleep to reinforce therapy is genuinely novel and being actively studied.

Wearable devices designed to interrupt nightmares in real time are also entering the picture. NightWare, an FDA-authorized prescription device worn as an Apple Watch app, monitors heart rate and movement during sleep and delivers gentle vibrations intended to disrupt a nightmare without fully waking you. A pilot study with first responders found that nightmare frequency dropped from about five nights per week to about three-and-a-half, and distress ratings fell significantly over 60 days.¹⁸ An earlier randomized trial of the same device in veterans showed both the active and sham groups improved, though the active device showed a stronger trend, and a post-hoc analysis restricted to frequent users found significantly better sleep quality.¹⁹ These results are promising but preliminary, and the sham-group improvement in the randomized trial highlights how powerful placebo effects can be in nightmare research.

Lucid Dreaming as a Nightmare Strategy

Lucid dreaming, the ability to recognize that you’re dreaming while you’re still asleep, is sometimes promoted as a way to take control of nightmares in the moment. The logic is appealing: if you know it’s a dream, you can change the plot, confront the threat, or simply decide to wake up. A systematic review found that most included studies did show lucid dreaming therapy to be effective in reducing nightmare frequency among adults with chronic and recurring nightmares.²⁰ Early case studies using progressive muscle relaxation, guided imagery, and lucid dream induction reported that four out of five subjects were nightmare-free at one-year follow-up.²¹

The catch is that lucid dreaming is genuinely difficult to learn. Most people can’t do it reliably even with practice, and the studies tend to be small. It’s probably best understood as a complement to IRT rather than a replacement. If you happen to be someone who can achieve lucidity during dreams, or you’re interested in training that skill, it may give you an additional tool. But if you’re looking for the most efficient path to fewer nightmares, IRT or ERRT will get you there with a much shallower learning curve.

Nightmares in Children

Kids have nightmares at high rates, especially between the ages of about three and six, and most outgrow them. When nightmares persist or are severe enough to cause sleep avoidance, fear of the dark, or daytime distress, treatment looks a bit different from the adult approach. Researchers have adapted ERRT for children with trauma-related nightmares, and preliminary case studies showed improvements in nightmare frequency, sleep disturbances, and parent-reported behavior problems.²²

A newer theoretical model designed specifically for youth nightmares, called DARC-NESS, emphasizes a concept the researchers call “nightmare self-efficacy,” which is basically the child’s belief that they can handle their nightmares. The model argues that this sense of mastery interacts with dream content, the child’s appraisal of how scary the dream was, their ability to self-regulate, conditioned arousal at bedtime, and overall sleep habits to either perpetuate or break the nightmare cycle.²³ In practical terms, helping a child feel that they have some power over their dreams (through rescripting exercises, drawing a different ending, or “choosing” a dream helper) appears to be a key ingredient. Parents should also look at basic sleep hygiene: consistent bedtimes, a cool and dark room, limited screen exposure before bed, and a calm pre-sleep routine go a long way when the nightmare problem is mild.

EMDR and Trauma-Focused Therapy

Eye Movement Desensitization and Reprocessing, or EMDR, is a well-established therapy for PTSD, and there is clinical evidence that it can also resolve trauma-related nightmares. In a case report using the standard eight-phase EMDR protocol delivered across three weekly sessions, the patient’s nightmare frequency dropped sharply within one week and nightmares ceased entirely afterward, with the improvement maintained at three-month follow-up.²⁴ EMDR works on the traumatic memories that fuel the nightmares rather than on the dream content itself, which makes it a better fit for people whose nightmares clearly stem from identifiable traumatic events. For idiopathic nightmares with no obvious trauma origin, IRT remains the more direct option.

Why the Brain Produces Nightmares in the First Place

One influential theory proposes that dreaming evolved as a threat-simulation system. The idea is that the brain rehearses dangerous scenarios during sleep to keep threat-detection skills sharp, much like a fire drill. Evidence supporting this comes from the fact that threatening content dominates normal dream reports across cultures, and children’s dreams are especially heavy on threats like being chased by animals.²⁵ Under this framework, nightmares are the system overshooting, the drill that becomes too realistic and wakes you up. This doesn’t mean nightmares are useful or that you should just tolerate them. A smoke alarm that goes off every time you boil water is still doing something related to its design purpose, but you’d be right to fix it. The threat-simulation hypothesis mainly helps explain why nightmares feel so viscerally real and why the most common nightmare themes, being chased, falling, losing control, are so universal. The treatments described above essentially retrain the system to run the drill at a lower intensity or with a less distressing script.

Interestingly, a large neuroimaging study attempted to find a clear brain-connectivity signature for nightmare frequency by looking at how the amygdala and prefrontal cortex communicate, and found no significant relationship in either of two independent large datasets.²⁶ This doesn’t mean the brain regions aren’t involved. It means that the relationship is probably more about how these areas behave during sleep itself, in moments of high emotional load, rather than something visible in a resting-state brain scan. Nightmare neuroscience is still catching up to the clinical treatments, which, for once, are actually ahead of the basic science.