Paxil is the SSRI (selective serotonin reuptake inhibitor, the standard first-line antidepressant class) I almost never start a new patient on, and the…
Sections
Paxil is the SSRI (selective serotonin reuptake inhibitor, the standard first-line antidepressant class) I almost never start a new patient on, and the SSRI I sometimes inherit patients on and then have a long conversation about whether and how to come off. It works. It’s been around since the early 1990s. It’s also got the worst withdrawal profile in the SSRI class, the highest anticholinergic load (which is a separate problem I’ll get to), and a side effect picture that, if I were picking an antidepressant for myself in 2026, would not be on the short list.
If we’re being honest, Paxil is the SSRI that got prescribed too aggressively in the 1990s and 2000s on the back of marketing that downplayed the discontinuation issues, and the medical literature has been quietly catching up to that for the last twenty years. Any prescriber who tells you Paxil withdrawal is mild is, at this point, a damn liar, or hasn’t been paying attention.
What it does
Paroxetine is an SSRI, and at higher doses it also hits norepinephrine reuptake and has some mild anticholinergic activity, meaning it blocks acetylcholine receptors in a way the other SSRIs mostly don’t. Acetylcholine is the neurotransmitter that runs a lot of the housekeeping functions, salivation, gut motility, urination, short-term memory, and blocking those receptors gives you the dry mouth, constipation, urinary retention, blurred vision, and (the part that matters most for older patients) cognitive effects you sometimes see on Paxil.
It’s FDA-approved for depression, generalized anxiety, social anxiety, panic, OCD (obsessive-compulsive disorder), PTSD (post-traumatic stress disorder), and PMDD (premenstrual dysphoric disorder). The breadth of approvals is most of why it got prescribed so heavily in the 1990s and 2000s. The clinical picture in 2026 looks different mostly because we have alternatives that don’t carry the same baggage, and the older alternatives that exist (Zoloft, Lexapro, Wellbutrin) mostly do what Paxil did, with less of the downside.
The withdrawal problem, which is the headline
Paxil has a short half-life (about 20 hours, but with nonlinear kinetics that mean the half-life shortens as the dose drops) and tight serotonin transporter binding. Coming off it is well-known to be brutal. The same brain zaps (electric-shock sensations in your head when you move your eyes), dizziness, nausea, flu-like symptoms, vivid dreams, and emotional rawness that show up in Effexor withdrawal show up in Paxil withdrawal, often just as severe, sometimes worse.
The studies on this are decades old. The genre of Paxil-withdrawal lawsuits and the Surviving Antidepressants subreddit are not making it up. The drug came onto the market with the marketing line that SSRIs don’t have withdrawal because they’re not addictive, which is true in the dependence-and-craving sense (you don’t crave Paxil the way you’d crave a stimulant) and entirely misleading in the practical “what does it feel like to stop” sense.
Tapering off Paxil is a months-long project for most patients. The taper goes slow, sometimes with liquid formulations to allow finer dose decrements at the bottom. Sometimes the move is to bridge to Prozac (fluoxetine) the same way as with Effexor, where Prozac’s long half-life essentially tapers itself once you stop it. The taper that gets through Paxil successfully is a six-to-twelve-month project for many patients, and that’s a real thing to know going in. Not to be Chicken Little about it, but a patient starting Paxil today is signing up for a potential six-month exit project, and that should be part of the conversation on day one.
The anticholinergic load and the rest of the side effects
Paxil is the SSRI with the most anticholinergic activity in the class. Practical consequences in the day to day: dry mouth, constipation, urinary retention (the kind of “I really need to pee but it won’t quite come” thing that older guys complain about), blurred vision, and the cognitive piece. There’s a literature on anticholinergic burden and dementia risk in older adults, and Paxil contributes more to that burden than the other SSRIs do. For a 32-year-old this is a footnote. For a 65-year-old it’s a real factor in the choice, the kind of factor that makes Paxil the wrong drug for that patient even if it would work for the depression.
Paxil also has higher rates of weight gain than most SSRIs, often a meaningful amount over a year or two. Twenty to thirty pounds isn’t unusual on a long Paxil course, which sneaks up on patients who notice their clothes fitting differently and don’t connect it to the medication they’ve been on since their forties. Sexual side effects are common and often persistent, the delayed orgasm and reduced libido picture that’s worse on Paxil than on most of the rest of the class. It’s also associated with more sedation than the other SSRIs in the class.
It’s category D in pregnancy due to cardiac malformation signal in early studies, which means it’s actively avoided in women who might become pregnant. Not super relevant to most of the guys taking it, but relevant to the wives in the picture and worth knowing for completeness, because the family-planning conversation around antidepressants is one most prescribers don’t think to have until somebody’s actively trying to conceive.
When (rarely) to use it
There are a couple of niches where Paxil still earns its keep. Social anxiety disorder is one, the data on paroxetine for social anxiety is among the strongest in the SSRI class, and a meaningful number of patients respond to Paxil for social anxiety after other SSRIs haven’t moved the needle. Premature ejaculation is another, where the sexual side effects everyone else hates are actually the therapeutic mechanism, and where Paxil sometimes gets used at low doses specifically for that purpose. PTSD where other agents have failed is a third.
For uncomplicated depression in a new patient, there’s almost no scenario in 2026 where Paxil is the right first pick. Lexapro, Zoloft, Celexa, Wellbutrin all do roughly what Paxil does without the same downside profile. The exception is patients who have a personal history of responding to Paxil in the past and not responding to other agents. The drug still works, the question is whether the cost is worth it for the patient in front of you, and the answer for new starts is usually no, because the alternatives are good enough.
The conversation when you’ve inherited a long-term Paxil patient
The patient I think about a lot is the guy who’s been on Paxil for fifteen or twenty years, started in his thirties for what was probably a real episode of panic or social anxiety at the time, and is now in his fifties on a drug that’s accumulated some baggage along the way. He’s gained thirty pounds. His sleep is worse than it should be. His constipation is bad enough that he’s quietly miserable about it but hasn’t connected it to the medication. His memory isn’t what it used to be, which he attributes to being in his fifties.
The Paxil hadn’t been wrong twenty years ago. The panic disorder was real and the medication worked. The cost-benefit just got worse as he aged and the alternatives got better. The conversation about whether to taper off is its own appointment, sometimes multiple appointments, and the plan that actually works has to be slow enough that he doesn’t hit the wall and quit halfway. Eight months. Liquid formulation for the bottom of the taper. Lexapro on standby in case the panic comes back. Most patients who do this carefully get all the way off and feel meaningfully better six months out, lighter, sharper, less bothered by the GI stuff. The panic doesn’t always come back. Sometimes it does, and the plan was to restart Lexapro instead of Paxil if it did, and that’s the version where the long-term patient ends up on a cleaner drug than they started on.
The patient autonomy piece
If you’re a long-term Paxil patient and you don’t want to taper off, the answer is still your call. Provider, not parent. The appointment isn’t mine. Some patients have built a working life around a drug that’s doing its job, and the discontinuation cost of trying to switch is real and not always worth taking on. Disapproving yes for the cases where I’d have picked a different drug if it were my first prescription, but yes regardless. I hardly ever say no. The honest take is what I’m here for, the choice is yours.
What I’m going to do at every visit is keep an eye on the trade-off. Weight check. A sense of how the GI piece is going. A real conversation about memory and sleep, not the rushed version. If the trade is getting worse and not better, that’s information that should change the conversation about whether to stay the course, not a verdict the prescriber gets to impose from the chair.
What’s nice to hear about Paxil specifically
For the patient where Paxil is actually the right drug, the social anxiety patient who didn’t respond to two other SSRIs and then came back to life on Paxil, the drug works in a way that the alternatives didn’t, and that’s worth saying. It’s a real medication and a real tool. The reputation it has now is mostly about the cases where it got used as a first-line drug for everything when it didn’t need to be, and the cases where the withdrawal got minimized by prescribers who should have known better. For the right indication and the right patient, Paxil earns its place.
The drug came onto the market with the marketing line that SSRIs don’t have withdrawal because they’re not addictive, which is true in the dependence sense and entirely misleading in the practical sense.
Bottom line
Paxil works. It also costs more than most of its alternatives in side effects, anticholinergic load, withdrawal severity, and weight gain. For most uncomplicated patients in 2026 there’s a better SSRI. For patients already on Paxil and doing well, the answer isn’t always to switch… if it’s working and tolerable and the patient wants to stay on it, the discontinuation cost of leaving may be higher than the marginal benefit of changing. But if it’s not working great, or the side effects have crept up over time, it’s worth a real conversation about whether the next twenty years should look the same as the last twenty.
Sources
- U.S. Food and Drug Administration. Paxil (paroxetine hydrochloride) Prescribing Information. NDA 020031. FDA; 2021. FDA label.
- Zhang B, Wang C, Cui L, et al. Short-Term Efficacy and Tolerability of Paroxetine Versus Placebo for Panic Disorder: A Meta-Analysis of Randomized Controlled Trials. Front Pharmacol. 2020;11:275. PMID 32296330.
- Li X, Hou Y, Su Y, et al. Efficacy and tolerability of paroxetine in adults with social anxiety disorder: A meta-analysis of randomized controlled trials. Medicine (Baltimore). 2020;99(14):e19573. PMID 32243377.
