Can Antidepressants Stop Working After Years?

The short answer: yes, depression can return while you still take a medication that worked for years. Reviews of published trials put the rate of returning symptoms during maintenance treatment at 9% to 57%, depending on the patients studied and the length of follow-up. Researchers do not agree on why. True drug tolerance is one theory among several, and it is not proven. Do not stop or change your dose on your own. Tell your prescriber, because several studied options exist.

Is antidepressant "poop-out" real?

The pattern is real. The cause is disputed. Clinicians call it tachyphylaxis, breakthrough depression, or loss of response. One review defines it as losing a previously effective response "despite staying on the same drug and dosage for maintenance treatment" (Targum, 2014, Innovations in Clinical Neuroscience).

The name suggests the drug wore off. That is an assumption. A 2019 review found that definitions of the problem "have also been largely inconsistent" across studies, and that it is unclear how much of it reflects a drug effect at all (Fornaro et al., 2019, Pharmacological Research).

This is different from a medication that never worked. Our guide to treatment-resistant depression covers that situation.

How often does depression come back while you stay on medication?

The numbers vary with the group studied.

• Reviews of published trials. A 1998 review reported a return of depressive symptoms during maintenance treatment in 9% to 57% of patients (Byrne and Rothschild, 1998, Journal of Clinical Psychiatry). A 2019 review reported the same range, 9% to 57%, "depending on the patient population and duration of follow-up" (Kinrys et al., 2019, Journal of Affective Disorders).
• Pooled randomized trials. A meta-analysis of 40 studies with 8,890 participants found a relapse rate of 20.9% in people who continued their antidepressant and 39.7% in people switched to placebo. In trials that lasted over one year, the rates were 19.9% and 39.8% (Kato et al., 2021, Molecular Psychiatry).
• Long-term users in primary care. The ANTLER trial enrolled 478 UK adults who felt well enough to consider stopping. Over 52 weeks, relapse occurred in 92 of 238 patients (39%) who stayed on their medication and in 135 of 240 (56%) who tapered to placebo (Lewis et al., 2021, New England Journal of Medicine).
• An observational study. In the NIMH Collaborative Depression Study, 103 people had 171 periods of maintenance medication after recovery. Depression recurred during 43 of those periods (25%). The median length of maintenance treatment was 20 weeks (Solomon et al., 2005, Journal of Clinical Psychiatry).

Staying on medication roughly halves the relapse rate in these trials. It does not protect everyone: about one in five people in the pooled trials relapsed on the active drug. Our article on how long to stay on antidepressants covers the duration question.

Why would an antidepressant stop working?

Researchers list many explanations. Tolerance is only one of them.

• The illness came back. Depression is usually a recurring illness. Proposed explanations include an increase in disease severity and "prophylactic inefficacy", which means the drug treats an episode but does not fully prevent the next one (Byrne and Rothschild, 1998).
• Missed doses. Reviews name medication nonadherence as a cause unrelated to tolerance (Targum, 2014). In ANTLER, adherence to the assigned treatment was 70% in the group that stayed on medication (Lewis et al., 2021).
• Bipolar depression that was not recognized. The reviews list unrecognized rapid cycling and "latent bipolarity" as possible explanations (Byrne and Rothschild, 1998; Fornaro et al., 2019). Our sister site bipolar.md covers how bipolar depression differs.
• Another medical condition. The VA/DoD guideline tells clinicians to rule out depression secondary to other causes, such as hypothyroidism, vitamin B-12 deficiency, pain, and chronic disease (VA/DoD Clinical Practice Guideline, 2022).
• Other medicines, alcohol, and drugs. MedlinePlus tells patients to report every medicine, herb, and supplement, and to "be honest about recreational drugs and alcohol, too" (MedlinePlus).
• Life stress. The 2019 review names "the intrusion of the impact of psychosocial variables into the clinical course" as one candidate (Fornaro et al., 2019).
• Loss of an early placebo effect. The 1998 review lists this as a possibility.
• True tolerance. The same review lists pharmacologic tolerance. The 2019 review points to possible adaptations in the stress hormone system and the serotonin system. These are hypotheses.

Is it a relapse or a side effect?

Sometimes the problem is not low mood. One researcher noted that after successful treatment, many patients report apathy or low motivation ("the blahs"), fatigue, dull thinking, sleep disturbance, weight gain, and sexual dysfunction. Scores on his seven-item scale for these complaints did not track with depression severity on a standard rating scale (Rothschild, 2008, Comprehensive Psychiatry).

Tell your prescriber which one you have: returned sadness and hopelessness, or flatness and fatigue with a steady mood.

What will a prescriber check first?

The VA/DoD guideline has no separate recommendation for a response that was lost after years. Its advice for a partial or absent response is the closest match. It says patients "should be reassessed for possible diagnostic error, the presence of co-occurring conditions, and treatment adherence" before treatment changes (VA/DoD Clinical Practice Guideline, 2022).

Those three checks suggest the questions to prepare for:

• How many doses you miss in a typical week.
• New prescriptions, over-the-counter medicines, and supplements.
• Alcohol and drug use.
• Any past period of unusually high energy, little need for sleep, or racing thoughts.
• New health problems, pain, and recent stress.

The guideline also states that "surveillance for recurrence or relapse should continue indefinitely."

What are the options if it stops working?

The evidence here is thin. The 2019 review found "few established treatment strategies" and only limited evidence for four approaches: a dose change, a switch of antidepressant class, augmentation or combination, and psychotherapy (Kinrys et al., 2019). An earlier review found no controlled trials on the question at that time (Nierenberg and Alpert, 2000).

• Dose increase. In one fluoxetine study, patients who relapsed on 20 mg daily had the dose raised to 40 mg daily, and 57% responded. Across the two dose-increase groups, 35% did not respond or relapsed again (Schmidt et al., 2002, Psychotherapy and Psychosomatics). The VA/DoD guideline cites a separate review that found no clinically relevant dose-effect gradient for SSRIs and more dropouts at higher doses.
• Switch. For a partial or absent response, the guideline suggests a switch to another antidepressant or to psychotherapy. See how doctors switch antidepressants for the methods.
• Augmentation. The guideline suggests two forms: psychotherapy added to the medication, or a second-generation antipsychotic.
• Added psychotherapy. In a pilot study of 10 patients who relapsed on medication, four of five responded to a larger dose, but all had relapsed again by one year. Four of five responded to cognitive behavior therapy at an unchanged dose, and one relapsed (Fava et al., 2002, American Journal of Psychiatry). A 20-patient pilot of a family intervention found a similar pattern (Fabbri et al., 2007). Both studies are small.

When should you call your prescriber?

Call when symptoms return and stay. MedlinePlus is direct: "you should never change your dose or stop taking an antidepressant on your own." It adds that stopping too fast may make depression come back (MedlinePlus).

Thoughts of suicide need help now. NIMH advises people to call or text the 988 Suicide and Crisis Lifeline at 988, or to chat at 988lifeline.org. In a life-threatening situation, call 911 (NIMH).

The bottom line

An antidepressant can lose its effect after years. In pooled trials, about 20% of people relapsed while they stayed on the active drug, against about 40% on placebo. The cause is often something other than tolerance: a new episode, missed doses, a new medicine or illness, alcohol, or bipolar depression that was missed. A dose change, a switch, augmentation, and added psychotherapy all have some evidence, but the studies are small. See your prescriber. Do not stop the medication yourself.

Last updated: October 2026. This article is for informational purposes only and does not constitute medical advice. Do not stop, start, or change the dose of an antidepressant without your prescriber. If you are in crisis, call or text 988.