What Happens When You Stop Taking Antidepressants?

Two things happen when you stop an antidepressant, and conflating them causes more harm than either one alone.

The first is discontinuation syndrome: a set of mostly physical symptoms caused by the brain adjusting to the drug's absence. It begins within days and usually fades within weeks. The second is relapse: the return of the depression itself, which typically develops over weeks and looks like the original illness.

People who mistake withdrawal for relapse conclude they need the medication permanently. People who mistake relapse for withdrawal wait it out until they are seriously unwell. Telling them apart is the single most useful skill here.

The short answer

Roughly one in six to seven people experience withdrawal symptoms directly caused by stopping an antidepressant, and about one in 35 experience severe ones. Symptoms usually begin within days and resolve within weeks. Separately, stopping at all roughly doubles the risk of relapse over the following 6 to 12 months, and that risk is reduced by gradual tapering combined with psychological support.

Neither figure is an argument for staying on medication forever, and neither is an argument for stopping abruptly. They are arguments for stopping deliberately, with a plan.

How common are withdrawal symptoms, really?

This has been genuinely contested. Estimates in circulation range from "rare and mild" to "more than half of patients," which is not a useful spread when you are trying to make a decision.

The most comprehensive answer comes from a systematic review and meta-analysis published in The Lancet Psychiatry in June 2024. It pooled 79 studies, 44 randomized controlled trials and 35 observational studies, covering 21,002 patients: 16,532 discontinuing an antidepressant and 4,470 discontinuing placebo.

The results, and why the placebo arm matters:

• 31% of people stopping an antidepressant reported at least one discontinuation symptom.
• 17% of people stopping placebo reported the same thing.
• The difference, roughly 15% or one in six to seven, represents symptoms attributable to stopping the drug itself.
• Severe symptoms occurred in about 3%, one in 35, with roughly 2% attributable to discontinuation after accounting for placebo.

That placebo figure is not a footnote. It means about half of the symptoms people report after stopping are not caused by the pharmacology, and are instead expectation effects, ordinary life, or the underlying condition. It also means the remaining half are real and should not be dismissed, which is what patients reporting withdrawal have been told for years.

The honest summary: withdrawal is real, it affects a substantial minority rather than a majority, and severe cases are uncommon but not vanishingly rare.

What do discontinuation symptoms feel like?

They are dominated by physical and sensory complaints rather than mood:

• Dizziness and imbalance. Often the first and most disruptive symptom. Frequently described as lurching or vertigo when turning the head.
• Flu-like symptoms. Fatigue, muscle aches, sweating, chills, headache.
• Nausea and other gastrointestinal upset.
• Sensory disturbances. Including brain zaps, brief electrical-shock sensations in the head often triggered by moving the eyes.
• Insomnia and vivid dreams.
• Irritability, agitation, and anxiety. This is the overlap zone with relapse and the reason the distinction gets confusing.

Onset is usually within two to four days of stopping or reducing, and often within 24 hours for short half-life drugs.

How do you tell withdrawal from relapse?

Three tests, in order of usefulness:

1. Timing. Withdrawal starts within days of a dose change. Relapse typically takes weeks to build. Symptoms that appeared 48 hours after your last dose are almost certainly withdrawal.

2. Symptom character. Withdrawal leads with the body: dizziness, nausea, electrical sensations, flu-like malaise. Relapse leads with mood and cognition: persistent low mood, loss of interest, hopelessness, worthlessness. Dizziness and brain zaps are not depression symptoms.

3. Response to reinstatement. This is the decisive test and it is the reason to involve your prescriber early. If the previous dose is restarted, withdrawal symptoms usually improve within hours to a couple of days. A depressive relapse does not respond that fast, because antidepressants take weeks to work in the first place, as anxiety.md covers here.

Rapid improvement after reinstatement is strong evidence for withdrawal. It also usefully rules out the conclusion many people draw, which is that they are simply unable to function without the drug.

Which antidepressants are hardest to stop?

Half-life is the main driver. Drugs that clear quickly produce a steeper fall in blood level, and a steeper fall produces more symptoms.

The 2024 meta-analysis found that stopping imipramine, paroxetine, and venlafaxine or desvenlafaxine was associated with a higher risk of severe symptoms than other antidepressants. All three have relatively short half-lives, which fits the mechanism.

At the other end, fluoxetine has an unusually long half-life, and its active metabolite persists for weeks. In effect it tapers itself, and discontinuation symptoms are correspondingly less common. This is why prescribers sometimes switch a patient from a short half-life drug to fluoxetine and then stop the fluoxetine, a strategy known as a bridge or switch taper.

None of this makes paroxetine or venlafaxine bad drugs. It means the exit plan needs to be slower.

How should a taper actually work?

The conventional advice, halving the dose every couple of weeks, is increasingly viewed as too fast at the low end. The reason is pharmacological: the relationship between dose and receptor occupancy is not linear. Going from 20 mg to 10 mg changes occupancy far less than going from 5 mg to zero. The last small step is often the hardest one.

Principles that reflect current practice:

• Reduce by proportion, not by fixed amounts. Cutting by roughly 10% to 25% of the current dose at each step, rather than a fixed number of milligrams, keeps the steps even in effect.
• Slow down as the dose gets lower. The final stretch usually needs the smallest and slowest steps.
• Hold when symptoms appear. Stay at the current dose until they settle rather than pushing through.
• Do not taper during a crisis. A stable period is a much better starting point than a stressful one.
• Ask about formulations. Liquid preparations, smaller tablet strengths, or tapering strips make small reductions practical. Splitting tablets is imprecise, and some formulations must not be split at all.

Timeframes vary widely. Some people come off comfortably in a few weeks. Others need many months. Neither is a sign of weakness or of a broken brain.

Does tapering protect against relapse?

Here the 2024 findings need a companion, because withdrawal and relapse are different questions and the answers point in different directions.

A 2025 network meta-analysis in The Lancet Psychiatry covering more than 17,000 adults compared abrupt stopping, rapid tapering, gradual reduction over at least four weeks, gradual reduction with psychological support, and continuing at full or reduced dose.

Two findings matter:

• Stopping antidepressants by any method, in the absence of psychotherapy, at least doubled the chance of becoming unwell again within 6 to 12 months.
• Gradual dose reduction over at least a month combined with psychological support offered protection against relapse comparable to staying on full-dose medication.

The practical implication is unambiguous and often missed: the therapy component is not an optional extra alongside the taper. In this data it is what makes stopping as safe as continuing. Planning a taper without also arranging psychological support is planning the harder version.

What if the symptoms don't stop?

Most discontinuation symptoms resolve within weeks. A minority of people experience a prolonged course, and that experience has historically been poorly acknowledged.

If symptoms are still present after several weeks, or are worsening rather than settling, reasonable next steps include reinstating the last tolerated dose and stabilizing before restarting a slower taper, switching to a longer half-life drug to taper from, and reviewing whether something else is contributing, such as thyroid disease, anemia, sleep apnea, or an inner-ear problem causing the dizziness.

Two symptoms are not something to wait out. New or worsening suicidal thoughts, and severe agitation, both warrant contacting your prescriber immediately. If you are in crisis, call or text 988.

When is the right time to stop?

The general clinical convention after a first depressive episode is to continue treatment for a period of months after full remission rather than stopping as soon as you feel better, because the early months after recovery carry the highest relapse risk. People with multiple prior episodes are typically advised to continue longer.

Factors that argue for waiting: a recent episode, several previous episodes, ongoing major stressors, an unstable period at work or home, or no therapy in place. Factors that argue it may be a reasonable time: sustained remission, a stable life period, psychological support arranged, and a prescriber who has agreed a schedule with you.

One reframe worth holding onto. Needing a slow taper is not evidence that the medication was a mistake, and it is not evidence of addiction. Physical dependence in this sense means the body adapted to a drug, which is also true of many blood pressure medications. It is a reason to stop carefully, not a reason for regret.

If your reason for stopping is that the medication is not working rather than that you no longer need it, that is a different problem with different options, covered in this guide to treatment-resistant depression. And if you are weighing whether to be on medication at all, depression.md compares the alternatives here.