The short answer: Stopping lithium raises the risk of a new mood episode, and the risk arrives fast. In a pooled analysis of 257 people with bipolar I disorder, more than half of new episodes came within 10 weeks of stopping. How you stop matters most. Median time to relapse was 4.0 months after abrupt discontinuation versus 20.0 months after a gradual taper. Suicide risk also climbs in the first year off lithium. Some reasons to stop are legitimate, but the exit needs a plan and a prescriber.

How fast does relapse happen after stopping lithium?

Faster than the untreated illness would predict. In 1991, Suppes and Baldessarini pooled 14 studies covering 257 patients with bipolar I disorder who stopped lithium after an average of 30 months. More than 50 percent of new episodes occurred within 10 weeks. Half of patients had relapsed 5.0 months after their last dose.

Mania came first. The time to a 25 percent recurrence of mania was 2.7 months, versus 14 months for depression. The authors concluded that early recurrence risk may exceed the natural course of the disorder. Stopping is not a return to baseline. It can be a step below it.

The pattern holds in younger patients. Among 37 adolescents stabilized on lithium, the 13 who stopped relapsed at nearly three times the rate of those who continued. Because mania tends to return first, learn the early signs of mania before you change anything.

Does stopping lithium raise suicide risk?

Yes, and the signal is strongest in the first year. The National Institute of Mental Health notes that lithium can decrease the risk of suicide. Removing it removes that protection.

Baldessarini, Tondo, and Hennen followed more than 300 patients with bipolar I or II disorder before, during, and after lithium. During treatment, suicidal acts per year fell 6.5-fold. In the first year off lithium, mood illness recurred in 67 percent of patients, and suicidal rates rose 20-fold. Rates were much lower after that first year. Deaths by suicide were 14 times more frequent after discontinuation than during treatment.

Two details matter. First, 89 percent of suicidal acts occurred during depressive (73 percent) or mixed (16 percent) states. Second, suicidal risk was 2.0-fold lower after slow than after rapid discontinuation. If you have stopped lithium and feel depression or agitation building, call or text 988, the Suicide and Crisis Lifeline.

Why does a slow taper matter?

Because the rate of stopping drives most of the early risk. In a pooled study of 161 patients who had taken lithium for about 4.2 years, researchers compared abrupt discontinuation (over 1 to 14 days) with gradual discontinuation (over 15 to 30 days). Median time to recurrence was 4.0 months after abrupt stopping and 20.0 months after gradual stopping. In the first year, 6.5 percent of the abrupt group fell ill each month, versus 2.3 percent of the gradual group. Only 1.8 percent of abrupt stoppers stayed well for three years. Among gradual stoppers, 37 percent did.

A replication in 78 patients found 14.0 months to relapse after gradual discontinuation versus 2.5 months after rapid discontinuation, in both bipolar I and II.

A 2022 analysis of 227 patients confirmed it again. Median time to a new episode was 3.50 months after rapid discontinuation and 10.6 months after a taper of two weeks or longer. Rapid stopping, stopping for medical reasons, and a bipolar I diagnosis each raised early recurrence risk. Years on lithium and blood level did not.

The American Psychiatric Association practice guideline states that it is preferable to slowly taper a medication rather than discontinue it abruptly, at a time with no major anticipated stressors. Note that "gradual" in these studies meant at least two weeks, and many clinicians taper more slowly. The schedule is a decision for you and your prescriber.

Does lithium still work if you restart it?

For most people, yes. Tondo and colleagues followed 86 patients through two separate periods of lithium treatment averaging 4.6 and 4.4 years. Episodes per year were 0.83 in the first period and 0.94 in the second, and time spent ill was 18.0 percent versus 24.2 percent. Neither difference was statistically significant.

A minority respond less well. A 2024 review of six cohort studies (403 patients) found that between 3.6 and 27.7 percent of patients did not respond when lithium was restarted, with an average of 17.3 percent. Non-response was linked to a faster taper, a longer gap off the drug, and a longer illness before lithium was first started. The authors still concluded that the vast majority respond when lithium is restarted.

When is stopping lithium the right call?

Sometimes. In a cohort of 873 lithium patients, 54 percent stopped at least once. Adverse effects drove 62 percent of those stops, psychiatric reasons 44 percent, and physical health 12 percent. The most common side effects behind a stop were diarrhea (13 percent), tremor (11 percent), excess thirst and urination (9 percent), and a rise in creatinine (9 percent). Men were twice as likely as women to stop without consulting a doctor first.

Three medical reasons come up most often, and each has a standard approach:

  • Kidney function. The FDA label for lithium carbonate describes a chronic tubulointerstitial nephropathy with long-term use and directs clinicians to assess kidney function before and during treatment. Lithium should not be used when creatinine clearance falls below 30 mL/min. A falling kidney number calls for a planned taper, not an overnight stop.
  • Thyroid function. The label states that lithium concentrates in the thyroid and can cause hypothyroidism, so thyroid function is checked at three months and then every 6 to 12 months. It adds that supplemental thyroid treatment may be used if hypothyroidism develops. An underactive thyroid is usually treated, not a reason to lose a working mood stabilizer. See thyroid.md for how hypothyroidism is managed.
  • Pregnancy planning. A 2017 cohort of 1.3 million Medicaid pregnancies found cardiac malformations in 2.41 percent of infants exposed to lithium in the first trimester versus 1.15 percent of unexposed infants. The risk was smaller than older reports claimed, and it rose with dose. The other side is relapse. In a meta-analysis of 4,023 patients, women with bipolar disorder who were medication-free during pregnancy relapsed postpartum 66 percent of the time, versus 23 percent for those who stayed on prophylactic medication. The label adds that lithium is reduced or stopped two to three days before delivery, then restarted after birth.

One caution applies to all three. In the 2022 analysis, stopping for a medical reason was itself a predictor of early relapse. Medically necessary stops need the most planning, not the least. Our overview of lithium monitoring and side effects covers the labs that catch these problems early.

What should you do if you have already stopped?

Tell your prescriber today, even if you feel fine. The NIMH advises against stopping medication without talking to your health care provider first. The same logic applies after the fact.

  1. Do not restart on your own at the old dose. The FDA label calls for a serum lithium level after 3 days of dosing, drawn 12 hours after the last dose. Kidney function and other medications may have changed.
  2. Watch the first weeks closely. In the pooled studies, the first off-lithium episode matched the polarity of the first lifetime episode in 70 to 84 percent of cases. If your illness began with mania, expect mania first.
  3. Treat depression or a mixed state as urgent. Those were the states in which 89 percent of suicidal acts occurred. Call or text 988 if you have thoughts of harming yourself.
  4. Ask about alternatives if side effects were the reason. Our mood stabilizers guide compares the other options.

The bottom line

Stopping lithium is a high-risk event, not a neutral one. Relapse tends to come within months, mania usually comes first, and suicide risk rises sharply in the first year. The rate of stopping changes those odds more than any other factor: a taper of at least two weeks stretched median time to relapse from 4 months to 20 months. Most people who restart lithium respond again, but about one in six do not respond as well. Kidney disease, thyroid problems, and pregnancy are real reasons to reconsider lithium, and each has a managed path. If you have already stopped, call your prescriber. Do not wait.

Last updated: September 2026. This article is for informational purposes only and does not constitute medical advice. Never stop, restart, or change the dose of lithium without talking to your prescriber.