Prescribing Prazosin

Why Prazosin Matters

Prazosin is one of psychiatry's best examples of drug repurposing. It arrived in the 1970s as an antihypertensive, a short-acting alpha-1 adrenergic antagonist that relaxes vascular smooth muscle. It was never a very good blood-pressure drug (it sits low on every hypertension algorithm), and it would have been forgotten except for one observation: patients with post-traumatic stress disorder who took it stopped having nightmares.

That observation, pursued largely by Murray Raskind and colleagues at the Seattle VA over two decades, turned a forgotten antihypertensive into a drug with randomized trials for PTSD-related nightmares and sleep disruption, and it does so cheaply, without abuse potential, without sedation in the usual sense, and without the metabolic and cognitive costs of the alternatives. For context, the 2018 American Academy of Sleep Medicine position paper lists prazosin among the medications that "may be used" for PTSD-associated nightmares, and names image rehearsal therapy as the one treatment it recommends.

The thesis of this guide: prazosin works by turning down the nocturnal noradrenergic surge that drives trauma nightmares and hyperarousal, but you use it best by starting low, titrating patiently to nightmare response, and respecting its one real hazard, first-dose orthostatic hypotension. Do those three things and you have a targeted, well-tolerated, non-stigmatizing tool for a symptom domain that most other drugs do not touch.

Be clear about what prazosin is not. It is not an antidepressant. It is not an anxiolytic in the SSRI/benzodiazepine sense. It does not treat the mood, avoidance, or negative-cognition clusters of PTSD. It is a symptom-targeted agent aimed at one thing: autonomic hyperarousal, expressed most visibly as nightmares and disturbed sleep. Understand it, dose it, and explain it to patients on those terms.

Part 1: Indications

FDA-Approved Uses

• Hypertension (a less-preferred, second- or third-line antihypertensive)
• Benign prostatic hyperplasia: symptomatic relief of urinary obstruction

Every psychiatric use of prazosin is off-label. There is no FDA indication in PTSD or any mental-health condition. This does not weaken the case; it reflects that the drug is a cheap generic with no sponsor to fund a registration trial. Document the off-label rationale anyway.

PTSD Nightmares and Hyperarousal

This is prazosin's home turf, and the strongest evidence base for any psychiatric application of the drug.

PTSD trauma nightmares and sleep disturbance: the core indication. Pooled analyses of placebo-controlled trials, including the one large negative trial (see Part 10), still find a significant benefit for nightmares: a standardized mean difference of 0.75 favoring prazosin across six trials and 429 patients (Reist 2021), and 0.64 across ten trials and 648 patients (Mendes 2025). Sleep outcomes also improved significantly in both, but the effect on overall PTSD symptoms is borderline (0.31, with a confidence interval that just excludes zero, in the first; 0.43 and not statistically significant in the second). Nightmares and sleep disturbance are the target symptom cluster:

• Recurrent trauma-content nightmares
• Disturbed awakenings, sleep terrors, night sweats
• Nocturnal autonomic surges (tachycardia, hyperventilation during sleep)
• Overall fragmented, non-restorative sleep

PTSD daytime hyperarousal. Beyond sleep, prazosin can blunt daytime autonomic hyperreactivity: irritability, the fight-or-flight response to trauma reminders, hypervigilance, scanning behavior, inability to tolerate crowds. This is often the symptom set that gets mislabeled. Patients pick up a diagnosis of generalized anxiety disorder because they are anxious in so many settings, or get called bipolar because the irritability looks like mania. Prazosin, dosed to cover the daytime, can address it directly.

Who Benefits Most

• Patients whose PTSD is dominated by nightmares and sleep disruption (sleep is foundational in PTSD; when it improves, daytime symptoms often follow)
• Both civilian and combat/military PTSD have been studied. The positive placebo-controlled trials include active-duty soldiers (Raskind 2013) and a small crossover trial of 13 civilian outpatients (Taylor 2008); the largest trial, in veterans, was negative (see Part 10).
• Patients with comorbid hypertension: prazosin also lowers blood pressure, so one drug can address both (coordinate with the PCP).
• Patients with comorbid obstructive sleep apnea: prazosin has no muscle-relaxant effect and is safe here, unlike benzodiazepines and z-drugs. (PTSD populations have strikingly high OSA rates, 40 to 75% in some series, often in young, thin patients, so keep suspicion high.)
• Patients who want to avoid stigma: many like that it is "just a blood-pressure pill."

A Secondary Use: Alcohol Use Disorder

Prazosin has moderate and still-accumulating supporting data in AUD, on the theory that an overactive noradrenergic system drives craving. In one 12-week RCT (n=80, PTSD excluded), prazosin titrated to 16 mg/day produced an 8.0 drinks/week reduction vs 1.5 on placebo (p=0.03) and fewer heavy-drinking days. A Phase II trial paired prazosin ER with cyproheptadine (which blocks serotonergic impulsivity) and found a moderate, dose-related effect (effect size 0.36 to 0.44).

That makes it a reasonable second-line option for AUD, most attractive when the drinking coexists with anxiety, insomnia, nightmares, PTSD, or hypertension, i.e., when one drug can do several jobs. It is not first-line AUD pharmacotherapy on its own.

Part 2: Workup and Candidacy

Prazosin asks far less of you before you start than a narrow-therapeutic-index drug does. There are no routine labs, no serum levels, and no ECG required for typical patients. The entire pre-start assessment centers on one number: blood pressure.

The Minimal Pre-Start Assessment

Candidacy Notes

• Comorbid hypertension is not a contraindication. Do not steer away from prazosin because a patient is already on BP meds; coordinate. Blood pressure should not decide who gets a trial: a secondary analysis of the 2013 soldier trial linked higher baseline standing systolic pressure to greater benefit, but a similar analysis of a trial in veterans with PTSD and alcohol dependence did not.
• Baseline hypotension / low-normal BP is a caution. Prazosin lowers blood pressure and can cause syncope and orthostatic hypotension (label). This does not forbid a trial, but it raises the caution: start at the lowest dose and titrate slowly.
• High fall risk (frail elderly, unsteady gait) warrants extra caution, the lowest starting dose, and the slowest titration, not necessarily avoidance.
• Active, unstabilized substance use predicts a poor trial; stabilize first where possible.

Part 3: How to Start and Dose

Two dosing errors are worth avoiding: starting too high, which causes the first-dose collapse that scares everyone off the drug, and stopping titration early in a patient who is tolerating the drug and still has nightmares.

Formulation

Use generic immediate-release prazosin. It is cheap and universally available. Its half-life is short, so daytime coverage requires a separate morning dose rather than one big dose lasting all day. An extended-release formulation has been studied (in the AUD combination work, in particular) but is not the standard tool; IR dosed appropriately does the job.

Starting Dose: Start Low, at Bedtime

• Start 1 mg at bedtime (QHS). This is not a rounding preference; it is a safety rule. Starting at 2 mg produces a markedly higher rate of falls and first-dose syncope. The PDR explicitly recommends the 1 mg start for exactly this reason.
• Give the first dose at bedtime, with the patient already lying down for the night, so that if orthostasis occurs it happens while they are recumbent and asleep rather than standing in a hallway.

Titration

• Increase by 1 to 2 mg every 4 to 7 days, slowing down or holding if dizziness appears.
• Titrate to a defined clinical endpoint, not a fixed milligram number. The goal is "zero sleep disturbance": no nightmares, no disturbed awakenings, no sleep terrors. Keep going until you hit that or until orthostasis limits you.
• See the patient often. This is a high-touch titration, closer in intensity to a lithium trial than to starting an SSRI: frequent contact, careful dose adjustment, weeks-to-months to full effect.

Typical Effective Range

• Across published studies, nighttime doses ranged from about 1 to 16 mg QHS, lower in civilian and older patients and higher in veterans.
• Means in the 2013 soldier trial were higher than in earlier studies: about 16 mg/night in men and 7 mg/night in women.
• Titrate to nightmare response, as the trials did. The 2013 soldier trial titrated over 6 weeks based on nightmare response, to a possible maximum of 20 mg at bedtime for men and 10 mg for women. If a patient is tolerating a low dose with partial benefit, keep titrating against the nightmares.

Part 4: Monitoring

Monitoring prazosin is light work: there is nothing to draw from a vein. What you follow is clinical and cardiovascular.

• Blood pressure and pulse at baseline and with each titration step, especially in the elderly or anyone on other antihypertensives. Check for a symptomatic orthostatic drop.
• Symptom check at every follow-up: dizziness, lightheadedness, near-syncope, falls, headache.
• Treatment response, tracked against the explicit goal: zero nightmares, disturbed awakenings, and sleep terrors. Partial improvement is a signal to keep titrating, not to declare victory.
• No routine labs, no serum levels, and no routine ECG for the typical patient.

Monitor more closely during active titration, in the elderly, and whenever another blood-pressure-lowering agent is added.

Part 5: Side Effects and How to Manage Them

Prazosin is generally well tolerated, and in the controlled trials the large majority of adverse events were mild to moderate. Nearly everything that matters flows from one mechanism, alpha-1 blockade lowering blood pressure, and nearly all of it is managed by starting low, going slow, and teaching the patient to stand up carefully.

Orthostatic Hypotension and Dizziness

This is the adverse effect. It is worst with the first dose and with each dose increase, and worse with faster titration or a starting dose above 1 mg. It presents as dizziness, lightheadedness, palpitations, and, at the extreme, syncope.

Management:

• 1 mg start, dosed at bedtime; titrate 1 to 2 mg every 4 to 7 days. Orthostasis is far easier to prevent than to rescue.
• Hold or step back down if symptomatic dizziness emerges; resume titration more slowly.
• Teach the falls-prevention routine explicitly: this single conversation dramatically reduces injuries.

Coordinate with the PCP about reducing other antihypertensives rather than avoiding prazosin.

"Drowsiness" and Why Prazosin Is Not a Sedative

Patients may report mild drowsiness, but prazosin is not meaningfully sedating, no more than placebo in the trials. Its sleep benefit does not come from CNS depression; it comes from removing the nocturnal noradrenergic hyperarousal that was fragmenting sleep and generating nightmares. That matters for two reasons: it is safe in sleep apnea (no muscle relaxation, no respiratory suppression), and you should set patient expectations accordingly; this is not a knockout pill.

Headache

A recognized alpha-1-blocker effect, generally mild and often transient. Usually managed with reassurance and time; slow the titration if bothersome.

Priapism: Rare but Counsel on It

As an alpha-1 antagonist, prazosin carries a rare risk of priapism. It is uncommon, but it is a urologic emergency: a sustained, painful erection lasting hours can cause permanent damage. Counsel male patients that a prolonged, painful erection requires immediate emergency care. Rare does not mean it can be skipped in the conversation.

Edema

Reported in some trials, the AUD combination work among them; generally mild.

Nasal Congestion

A predictable consequence of alpha-1 blockade on nasal vasculature; usually a minor nuisance.

Part 6: Overdose and Toxicity

Prazosin has no boxed warning and a relatively benign overdose profile compared with many psychotropics, but the mechanism predicts exactly what an overdose looks like: exaggerated alpha-1 blockade leading to profound hypotension and syncope, with reflex tachycardia.

• Presentation: marked hypotension, dizziness, syncope, drowsiness; potentially shock at large ingestions.
• Management is supportive; the cornerstone is restoring volume and blood pressure: keep the patient supine, elevate the legs, give IV fluids, and use vasopressors if hypotension is refractory. This is an ED evaluation.
• Because the half-life is short, the hypotensive effect of an isolated overdose is generally time-limited once the patient is supported through it.

Part 7: Drug Interactions

Prazosin has minimal CYP450 drama. Almost every clinically important interaction is pharmacodynamic: additive blood-pressure lowering.

The Interactions That Matter: Additive Hypotension

• Other antihypertensives (any class): additive BP lowering. The right move is usually not to avoid prazosin but to coordinate with the PCP about trimming the other agents.
• PDE5 inhibitors (sildenafil, tadalafil, vardenafil): a clinically significant interaction. The combination can produce substantial, symptomatic hypotension. Counsel patients explicitly: separate the timing, use the lowest PDE5 dose, and warn them not to add an erectile-dysfunction drug on their own without discussing it.
• Alcohol: additive hypotension and dizziness; relevant given the AUD overlap.
• SSRIs: possible additive hypotensive effect; monitor BP, though this is a minor concern relative to the others.

Benzodiazepines

A 2025 meta-analysis raised the possibility that benzodiazepines may enhance prazosin's effect when co-administered. Treat that as hypothesis-generating and in need of prospective confirmation, not as a reason to add a benzodiazepine, which carries its own liabilities in PTSD: abuse risk, interference with exposure therapy. File it as "interesting," not "actionable."

What Prazosin Does Not Meaningfully Interact With

• No significant CYP450-mediated interactions of concern
• Compatible with most psychiatric medications; the recurring theme is simply additive hypotension, not pharmacokinetic collision

Part 8: Special Populations

The Elderly

This is the population where prazosin's one hazard bites hardest. Orthostatic hypotension plus an older patient equals falls and fractures.

• Start 1 mg at bedtime; titrate especially slowly.
• Counsel the falls-prevention routine and mean it.
• Monitor BP and standing tolerance more closely.
• The upside remains: comorbid hypertension is common in this group, and prazosin also lowers blood pressure.

Pregnancy and Lactation

Data are limited. Prazosin has been used to manage hypertension in pregnancy, and there is no established major teratogenic signal, but the psychiatric evidence base does not extend here with any strength. Use only when the benefit clearly justifies the limited safety data, and coordinate with obstetrics. For lactation, information is sparse; make an individualized decision and monitor the infant if used.

Renal and Hepatic Impairment

Prazosin is hepatically metabolized. Hepatic impairment may increase exposure; titrate cautiously. The distilled source data on renal/hepatic dosing are thin; the safe posture is to start low and go slow in organ impairment and lean on the same orthostasis precautions that govern everyone.

Cardiac Disease

Because prazosin lowers blood pressure and can provoke reflex tachycardia, use caution in patients with significant hypotension-sensitive conditions or those already running low pressures. In patients with heart failure it has been used for afterload reduction, but for psychiatric prescribing what matters is the baseline BP and the cardiac history: know both, and titrate accordingly.

Active Substance Use

Unstabilized substance use predicts a poorer trial and complicates adherence and interpretation. Stabilize where you can before judging efficacy.

Part 9: Discontinuation

Prazosin does not produce the dangerous rebound syndrome that defines lithium or benzodiazepine withdrawal. There is no significant physiologic dependence and no protracted withdrawal syndrome. Two caveats, though:

1. Rebound hypertension/tachycardia can occur with abrupt cessation of any alpha-blocker, particularly at higher doses, so in a patient at a substantial dose, taper rather than stop cold (e.g., step down by 1 to 2 mg every 5 to 7 days), mirroring the titration up.
2. The symptoms come back. Prazosin is symptomatic, not disease-modifying; it suppresses nightmares and hyperarousal while present. When it is stopped, the nightmares and disturbed sleep typically return. This is not withdrawal; it is the untreated illness re-emerging. Frame it that way for patients so a return of nightmares off the drug is understood as "we stopped the treatment," not "the treatment broke me."

Part 10: Prazosin vs the Other PTSD Nightmare Approaches

Trauma-focused psychotherapy remains first-line for PTSD overall. Prazosin is not a substitute for it; it is a targeted adjunct for the hyperarousal/nightmare domain, best used alongside evidence-based therapy. Nor is it a broad PTSD treatment: it does not address avoidance, mood, or negative cognitions.

Within the narrower question of nightmares and disturbed sleep, here is how it compares:

• vs SSRIs (sertraline, paroxetine, etc.): SSRIs are FDA-approved for PTSD but carry small effect sizes and famously underperform in combat-related PTSD in male veterans; they can also worsen sleep early on. Prazosin targets a different domain (hyperarousal/nightmares). The two are complementary and often combined, and when sleep is the presenting complaint, there is a good argument to start prazosin before the SSRI.
• vs trazodone: Trazodone helps patients fall asleep but, per experienced VA prescribers, does not improve nightmares or disturbed awakenings the way prazosin does; tolerance develops and cognitive fog can carry into the day. Reasonable to add for sleep onset, but it is not a nightmare drug.
• vs benzodiazepines: A poor choice in PTSD. One small trial showed no benefit on core symptoms; they carry high abuse risk in a population with 40 to 75% comorbid SUD, they impair the emotional processing that exposure therapy depends on, and they worsen sleep apnea. Prazosin has none of these liabilities.
• vs hydroxyzine: Modest benefit over placebo, inferior to prazosin head-to-head, but less hypotensive: a sensible fallback if prazosin's orthostasis is prohibitive. Watch the QT above 100 mg/day.
• vs clonidine (alpha-2 agonist): The usual second choice if prazosin fails. Impressive VA case series but zero controlled trials, and it is more likely to cause hypotension than prazosin, so if you switched because of hypotension, go to hydroxyzine, not clonidine.
• vs doxazosin (long-acting alpha-1 blocker): Attractive on paper (longer half-life, once-daily) but it penetrates the brain poorly, and two recent RCTs (2023, 2025) found little benefit. Now generally avoided.
• vs topiramate/quetiapine: Both have some nightmare data (topiramate 25 to 100 mg QHS; low-dose quetiapine 25 mg QHS) but carry their own burdens and orthostasis warnings. Prazosin is generally preferred first among the pharmacologic options.

How to Interpret the Big Negative Trial

You will hear that "a large trial showed prazosin doesn't work." That trial, Raskind et al, NEJM 2018 (VA Cooperative Studies Program #563), randomized 304 veterans with chronic PTSD and frequent nightmares to prazosin or placebo for 26 weeks. It did fail its primary endpoints: at 10 weeks there was no significant difference in the nightmare item of the CAPS, in sleep quality on the Pittsburgh Sleep Quality Index, or in the Clinical Global Impression of Change, and none at 26 weeks. Three points matter when you weigh it:

• Under-dosing does not explain it. Doses were escalated over 5 weeks to a daily maximum of 20 mg in men and 12 mg in women, and the average maintenance dose of prazosin was 14.8 mg.
• The authors suggested selection bias as a possible explanation: recruitment of patients who were mainly in clinically stable condition, in whom symptoms were thought less likely to respond to antiadrenergic treatment.
• A meta-analysis that included it (Reist 2021) found that prazosin reduced outcome measures in this trial about as much as in past studies, but with a relatively large placebo effect, particularly for nightmares.

Pooled analyses that include this trial still find a significant benefit for nightmares, but only a borderline benefit for overall PTSD symptoms (effect sizes in the indications section above).

Mechanism

PTSD involves a chronically over-driven central noradrenergic system. During sleep, normally a period of falling adrenergic tone, trauma-affected brains generate surges of norepinephrine that fragment sleep architecture and appear to drive the vivid, terrifying, trauma-content nightmares and the nocturnal autonomic storms (racing heart, sweats, gasping awakenings).

Prazosin blocks the alpha-1 adrenergic receptor. By antagonizing alpha-1 signaling, including in the CNS, which prazosin reaches (unlike the poorly-penetrant doxazosin), it dampens that noradrenergic hyperarousal at its receptor target. The result is quieter sleep, fewer nightmares, and, when dosed into the day, reduced daytime fight-or-flight reactivity.

The Bedside Cheat Sheet