Hydroxyzine

Hydroxyzine is one of those drugs that’s been sitting in the pharmacy for sixty years doing useful work, and most clinicians forget it exists until somebody reminds them. It’s an antihistamine. Same H1 receptor class as Benadryl, just older and a bit more selective. Decades before anyone marketed it as an anxiolytic, doctors noticed that patients on it for hives or pre-op sedation got noticeably calmer. Eventually the FDA gave it a formal anxiety indication, which it still has, and which most prescribers in 2026 act like they’ve never heard of.

The reason it matters now is simple. Benzodiazepines are Schedule IV, they cause tolerance and physiologic dependence, they kill people in combination with opioids and alcohol, and the regulatory environment has gotten increasingly hostile to prescribing them long-term. So a patient walks in with situational anxiety, panic spikes, sleep onset that won’t quit, and the prescriber needs something that works PRN. Hydroxyzine is sitting right there. No DEA schedule. No abuse liability. No respiratory depression. Onset around thirty minutes, half-life around twenty hours, and it actually works for somatic anxiety.

The catch, because there’s always a catch, is that it makes you sleepy. Sometimes pleasantly sleepy. Sometimes the kind of sleepy where you shouldn’t drive to your kid’s school. And it’s anticholinergic, which matters a lot if you’re 78 and already on three other things that dry your mouth and cloud your thinking.

Why most psychiatrists ignore it

Honest answer is laziness plus habit. Klonopin works in twenty minutes, patients feel it work, they thank you, the visit’s over. Hydroxyzine works less dramatically. It takes the edge off rather than removing the floor entirely. Patients who’ve been on benzos sometimes try it and report back that it “didn’t do anything,” by which they mean it didn’t feel like a benzo. The somatic load, the chest tightness, the racing heart, the keyed-up-can’t-sit-still feeling, those often come down on hydroxyzine. The cognitive sense of being chemically tranquilized doesn’t. For a lot of patients that ends up being exactly what they wanted once they stop expecting a benzo. They can take it and still function.

The other reason it gets ignored is that residency training emphasizes the new stuff. SGAs for everything, SSRIs and SNRIs for anxiety, gabapentinoids when you’re feeling adventurous. Older drugs get a paragraph. Hydroxyzine gets half a paragraph and a footnote about Vistaril for itching. By the time you’re three years out of training, you’ve forgotten it exists unless somebody on your team uses it routinely.

I had a woman come in last spring, mid-30s, GAD with panic features. She’d been on Klonopin 0.5mg twice daily for four years. Her prior prescriber retired and the new guy refused to continue it, which is its own conversation, but she was now in a forced taper she hadn’t agreed to and showing up in my clinic terrified. We slowed the taper down considerably, added hydroxyzine 50mg up to four times daily as needed, and started Lexapro underneath. Six months in she’s off Klonopin entirely, takes hydroxyzine maybe twice a week when something flares, and is the most surprised person in the room. Her words were something like, “I genuinely did not believe an over-the-counter feeling pill was going to do this.” It isn’t over-the-counter, and the dismissive read on it is wrong in ways that took her six months to find out.

The good uses, the bad fits, the dosing

Good for: generalized somatic anxiety, the chronic keyed-up baseline, anticipatory anxiety before specific events, sleep-onset insomnia when anxiety is the driver, itching from anxious skin-picking, and acute agitation when you don’t want to use a benzo. Reasonable for acute panic if the patient catches it early. Less reasonable if they’re already in a full panic attack and want something that works in eight minutes, because hydroxyzine won’t.

Bad for: anybody who needs immediate-onset relief during a peak episode, anybody with significant anticholinergic burden already (Parkinson’s patients, dementia patients, people on tricyclics, people on bladder anticholinergics), anybody who has to operate machinery or drive for work in the next several hours, and anybody who’s already failed it at a reasonable dose. There’s a real efficacy ceiling. Some patients need a benzo and there’s no clever way around it.

It won’t feel like a benzo. That’s the point. If your patient wants to feel like they took a benzo, hydroxyzine will disappoint them every time.

The dosing piece is where prescribers undershoot constantly. Twenty-five milligrams of hydroxyzine in an adult with real anxiety is basically homeopathic. The studies that showed efficacy against placebo and against buspirone used 50 milligrams four times daily. In practice I start most adults at 25-50mg PRN, tell them they can repeat in an hour if they need to, and let them work up to where it actually does something. Ceiling is around 400mg/day on paper, though I rarely go above 200.

The anticholinergic problem nobody mentions

Hydroxyzine blocks muscarinic receptors in addition to H1. That’s where dry mouth, urinary retention, blurry near vision, and constipation come from. In a 28-year-old with panic disorder this is mostly a nuisance. In a 75-year-old with mild cognitive impairment, it’s a one-way ticket to a delirium consult. The American Geriatrics Society Beers Criteria flags it, and they’re right to. Cumulative anticholinergic burden over years is associated with dementia risk in observational data. The signal isn’t huge, but it’s there, and stacking hydroxyzine onto an oxybutynin onto a tricyclic onto a diphenhydramine sleep aid is exactly how you get a confused grandmother who used to be sharp.

For older patients I either skip it entirely or cap it hard. 10mg at bedtime for sleep, maybe 25mg PRN with a clear plan to reassess in a month. If they’re on anything else with anticholinergic activity, the answer is usually a different drug. Buspirone, low-dose mirtazapine, or just being honest with the family that anxiety in late life is a different beast and trying to make it disappear with a pill is usually how things get worse.

PRN versus scheduled, and how to actually use it

PRN is the default and the easiest sell. Patient has predictable triggers, takes hydroxyzine an hour ahead, gets through the meeting or the flight or the funeral. Less useful when anxiety is constant and there’s no clean trigger to time the dose against. For those patients, scheduled 50mg three or four times a day for a few weeks while you wait for an SSRI to work is reasonable. The somatic baseline drops. They sleep better at night because they’re less wound up during the day. Once the SSRI is on board you can usually peel the hydroxyzine back to PRN.

One thing I tell patients explicitly. This drug doesn’t build tolerance the way a benzo does. If 50mg worked in March and 50mg isn’t working in November, something else changed. Either the anxiety is worse, or there’s a new stressor, or sleep has fallen apart, or they’re drinking more, or the SSRI stopped working. The dose is almost never the variable that moved.

The post-script most prescribers should hear, and I include myself in this on bad days, is that hydroxyzine gets undervalued mostly because we’ve gotten used to thinking about anxiety medications in two categories. Daily preventive (SSRIs, SNRIs, buspirone) and rescue (benzodiazepines). Hydroxyzine is a third category that doesn’t get its own box. Rescue without the schedule and without the dependence, at the cost of sedation and a less dramatic subjective effect. For a meaningful slice of patients, that’s exactly the trade they’d take if it were offered. Most of them never get offered.