Prazosin for Nightmares

Prazosin is an alpha-1 adrenoreceptor antagonist. The FDA label carries exactly one indication, hypertension, and nothing else. Every use of it for nightmares is off-label prescribing, which is legal and routine but means the label will not give you a dose, a timeline, or an expected result. What moved it into psychiatry was an observation about noradrenaline: excessive brain responsiveness to norepinephrine appears to contribute to PTSD, particularly at night. Prazosin is brain-active, so it can blunt that signal where the signal is happening. The timing detail matters more than it sounds. Blood levels peak at about three hours and the plasma half-life is two to three hours, so a dose taken at bedtime works hardest across the early and middle part of the night and has largely worn off by morning.

That short window is why the drug is taken at night rather than with breakfast, and it is also why some trials added a smaller daytime dose when nightmares were not the only symptom being targeted.

Two randomized trials found a benefit. A 2007 parallel-group study in 40 veterans, averaging 13.3 mg a day over eight weeks, reported prazosin superior to placebo for trauma nightmares, sleep quality and global clinical status, with large effect sizes. A 2013 trial in 67 active-duty soldiers back from Iraq and Afghanistan, run across 15 weeks, reported benefit for nightmares, sleep quality and global function, at mean bedtime doses of 15.6 mg for men and 7.0 mg for women.

Then the largest trial landed and went the other way. The VA Cooperative Studies PACT trial randomized 304 veterans across 13 VA medical centers and found no significant difference from placebo at 10 weeks on distressing dreams (between-group difference 0.2; 95% CI -0.3 to 0.8; P=0.38), sleep quality (0.1; 95% CI -0.9 to 1.1; P=0.80) or global change (0; 95% CI -0.3 to 0.3; P=0.96). There was no difference at 26 weeks either. That is a large, well-run negative result on exactly the question this page asks.

Pooled analyses that include the null trial still come out in favor, narrowly, and only for the nightmares. A six-trial meta-analysis covering 429 participants found a significant benefit for nightmares (SMD -0.75; 95% CI -1.24 to -0.27), and its authors note the largest trial produced a reduction similar to earlier studies while its placebo arm improved so much that the gap closed. An eight-trial analysis found the same shape: nightmares improved (SMD -1.13; 95% CI -1.91 to -0.36) while overall PTSD symptoms and sleep quality did not reach significance.

In 2010 the American Academy of Sleep Medicine recommended prazosin for PTSD-associated nightmares at Level A, its strongest grade. Its 2018 position paper moved prazosin down into the may-be-used tier, which the paper defines as the position taken when the evidence or expert consensus is less clear. Check the methods before reading that as a verdict on the negative trial: the 2018 literature search ran from March 2009 to August 2017, and the VA trial published in February 2018. The downgrade arrived first and for other reasons.

The VA and Department of Defense moved in the opposite direction. Their 2017 guideline found insufficient evidence to recommend for or against prazosin for nightmares. The 2023 revision added a Weak for suggestion, stating that recent systematic reviews found a benefit even when the large VA Cooperative trial is included, while recording the work group's own confidence in the evidence as low because of small samples, heterogeneity and risk of bias. The two bodies are not looking at different data. They are weighting one placebo-heavy null result differently.

The distinction is written into the guideline itself. The same VA/DoD document that suggests prazosin for nightmares also suggests against prazosin for global PTSD symptoms, alone or added to something else. The pooled data line up with that split: the eight-trial analysis found improvement in nightmares but no significant effect on overall PTSD symptoms. If the hope is that a bedtime capsule will reach the daytime symptoms, the intrusive memories, the avoidance, the constant scanning, the evidence does not support that expectation.

What the guideline puts first is therapy. VA/DoD gives a Strong for recommendation to one-to-one manualized trauma-focused psychotherapies, Cognitive Processing Therapy, EMDR or Prolonged Exposure, and recommends those over medication for PTSD. Prazosin belongs beside that work. It does not replace it. Sleeping through more of the night is not a small thing when you are trying to do difficult therapeutic work in the daytime, and that is the honest case for it: a support for the treatment rather than the treatment.

Every trial named on this page enrolled a military population: veterans recruited at 13 VA medical centers, combat veterans with chronic PTSD and distressing trauma nightmares, active-duty soldiers returned from Iraq and Afghanistan. That is who the numbers describe. Whether the same effect shows up after a car crash, an assault or a medical trauma is a reasonable clinical extrapolation rather than a demonstrated one, and a clinician deciding suitability should say that out loud instead of presenting veteran data as universal.

The AASM does list prazosin as an option that may be used for nightmare disorder as well as for PTSD-associated nightmares, noting nightmare disorder affects roughly 4% of adults. The practical filters are cardiovascular. If you already take something for blood pressure, a diuretic, or a PDE5 inhibitor for erectile dysfunction, the VA/DoD guideline flags an additive hypotensive effect, and that combination needs deliberate handling rather than a checkbox on an intake form.

The AASM position paper recommends image rehearsal therapy for PTSD-associated nightmares and for nightmare disorder, the only treatment it places at that level in either category. Prazosin sits a tier below, on a long may-be-used list that also holds cognitive behavioral therapy, CBT for insomnia, EMDR, clonidine, cyproheptadine, fluvoxamine, gabapentin, nabilone, phenelzine, topiramate, trazodone, tricyclics and three atypical antipsychotics. Read that list as a map of how thin the evidence is across this whole problem. It is not a ranking.

Two entries are marked the other way: clonazepam and venlafaxine are not recommended for nightmare disorder. The guideline bodies disagree here too. The VA/DoD revision found insufficient evidence to recommend for or against image rehearsal therapy, exposure relaxation and rescripting therapy, or the NightWare device, which is the therapy AASM ranks highest. Doxazosin, a longer-acting drug in the same class, gets floated as an alternative but rests on a pilot crossover trial of eight enrollees.

The starting dose is not a judgment call and it is printed on the label: always start on the 1 mg capsules, and the 2 mg and 5 mg capsules are not indicated for initial therapy. For nightmares, the VA/DoD medication table lists 1 mg at bedtime, titrated to 3 to 20 mg at bedtime to clinical response. That is a different schedule from the label's hypertension dosing, which is 1 mg two or three times a day building to 20 mg in divided doses, so do not read one across to the other.

Titration is slow on purpose and it takes real calendar time. The 2007 trial reached an average of 13.3 mg a day over eight weeks. The 2013 soldier trial titrated over six weeks to mean bedtime doses of 15.6 mg for men and 7.0 mg for women. The PACT trial escalated over five weeks to a daily ceiling of 20 mg for men and 12 mg for women, in divided doses rather than a single bedtime dose. If a prescriber hands you a target dose on day one or wants a verdict after a week, that is not how any of these trials ran.

The label is blunt about this. Prazosin may cause syncope with sudden loss of consciousness, usually within 30 to 90 minutes of the initial dose, at an incidence of approximately 1% when that dose is 2 mg or greater. The same section says those episodes can be minimized by holding the first dose to 1 mg and increasing slowly afterward.

That one sentence is the entire reason titration is cautious. It describes a dose-timing problem with a known fix rather than a reason to avoid the drug.

More often the effect is milder: dizziness and lightheadedness, worse on rising from lying or sitting, with the label advising against driving or hazardous tasks for the first 24 hours and after any dose increase. In the PACT trial the supine systolic blood pressure fell 6.7 mm Hg further than placebo by 10 weeks. Prolonged erection and intraoperative floppy iris syndrome during cataract surgery both appear on the label, so an eye surgeon needs to know you take it.

One number is worth stating plainly, because this subject invites assumptions. In the PACT trial, new or worsening suicidal ideation was recorded in 8% of the prazosin group and 15% of the placebo group. If thoughts of harming yourself surface while you are being treated for trauma, that is a reason to contact your clinician promptly or reach the 988 Suicide and Crisis Lifeline. It is not something to hold until the next scheduled appointment.

Prazosin is a generic oral capsule and not a controlled substance, so a clinician licensed in your state can prescribe it after a telehealth visit across most of the country, and it fills at an ordinary pharmacy. Off-label prescribing is standard practice. The label's only indication is hypertension, and that does not stop a clinician from prescribing it for nightmares. What it does mean is that no regulator has vetted a nightmare dose, so the prescriber's own judgment carries more weight than usual.

The intake should reflect that. A service handling this properly asks what else lowers your blood pressure, whether you use a PDE5 inhibitor, whether you have fainted before, and what your trauma care looks like right now. It should start at 1 mg, put contact points in the calendar during titration, and refer for trauma-focused therapy rather than treat a capsule as the whole plan. A questionnaire that ships a starting dose above 1 mg, or leaves nobody reachable through six weeks of dose changes, is not built for this drug.

This is one of the cheapest drugs in the whole mental health formulary. Federal NADAC pricing published on 2 September 2026 lists generic prazosin at $0.08582 per 1 mg capsule, $0.08498 per 2 mg and $0.12929 per 5 mg, which puts a 30-capsule bottle of the starting strength near $2.57. Read that as what a pharmacy pays to acquire the drug. It is not a shelf price. The VA/DoD guideline describes prazosin as inexpensive and cites that as one reason it is easy to prescribe across settings.

We are not quoting a monthly telehealth price here, because the figure that reaches you is built from a consult fee, a dispensing fee and a margin that differ by service, and any number invented to fill the gap would be a guess dressed as data. Nothing here is medical advice, and the decision to try prazosin belongs to you and a prescriber who knows your history.

Ask instead whether the consult fee is separate from the medication, whether you can use your own pharmacy, who you can reach at 11pm during titration, and what happens if eight weeks of climbing changes nothing.