Can a manic episode last less than a week?

Under DSM-5, a manic episode needs 7 consecutive days unless the person is hospitalized. If hospitalization is needed, any length counts as mania. An elevated period of 4 to 6 days without hospitalization is classed as hypomania instead.

Will a bipolar episode end on its own without treatment?

Many episodes do end over months, and the long-term studies show most people recover within a year. But the studies followed people who could seek any care they wanted, so they do not show what happens with zero treatment. Mania in particular can cause serious harm before it ends, so waiting it out is not a safe plan.

Do bipolar II episodes last as long as bipolar I episodes?

The 25-year duration study covered bipolar I only. In that group, hypomania was the shortest episode type, with a median of 3 weeks. Bipolar II involves hypomania plus major depression, and the depressive phases tend to be the part that lasts.

What is a cycling episode?

A cycling episode is one continuous period of illness where the person moves between mania or hypomania and depression without 8 weeks of wellness in between. Researchers count it as a single long episode. In the NIMH study these were the longest episodes by far, with medians of 42 to 61 weeks.

How Long Does a Bipolar Episode Usually Last?

Q: How do doctors decide that an episode is over?

Research studies use a fixed rule. In the NIMH Collaborative Depression Study, recovery meant at least 8 consecutive weeks with no symptoms, or only 1 or 2 mild symptoms and no loss of function. A few good days do not count as the end of an episode.

The short answer: There is no single number, but there are good estimates. Diagnostic rules set minimums: mania lasts at least 7 days, hypomania at least 4 days, and a depressive episode at least 2 weeks. In a 25-year NIMH follow-up study of bipolar I disorder, the median episode lasted 13 weeks. Depressive episodes ran longest (median 15 weeks), manic episodes shorter (7 weeks), and hypomanic episodes shortest (3 weeks).

What counts as a bipolar episode?

The DSM-5 sets minimum durations for each type of mood episode. Clinicians use these floors to decide whether a mood shift is an episode at all. According to the StatPearls summary on the NCBI Bookshelf, the rules are:

Manic episode: elevated or irritable mood with increased energy for at least 7 consecutive days, or any length if the person needs hospitalization.
Hypomanic episode: the same kind of mood change for at least 4 consecutive days, without hospitalization or psychosis.
Major depressive episode: 5 or more depressive symptoms nearly every day for a consecutive 2-week period.

The National Institute of Mental Health describes bipolar I as "manic episodes that last at least 7 days (most of the day, nearly every day)," with depressive episodes "typically lasting at least 2 weeks." MedlinePlus puts it simply: "Your mood episodes may last a week or two or sometimes longer. During an episode, symptoms usually occur every day for most of the day."

These numbers are thresholds, not averages. They tell you the shortest stretch that qualifies. Most real episodes run well past them.

How long do episodes last in real life?

The best data come from the NIMH Collaborative Depression Study. Solomon and colleagues published the episode-duration results in Archives of General Psychiatry in 2010. The study enrolled 219 people with bipolar I disorder between 1978 and 1981 and followed them for up to 25 years (mean 17.3 years, median 20 years).

The headline finding: the median duration of a bipolar I mood episode was 13 weeks. More than 75% of participants recovered from their episodes within 1 year of onset. Of the 219 subjects, 79% recovered from their first observed episode within a year.

The full text on PubMed Central breaks the medians down by episode type:

Major depression: 15.0 weeks
Mania: 7.0 weeks
Minor depression: 7.0 weeks
Hypomania: 3.0 weeks

Compared with a major depressive episode, recovery from mania was 71% more likely at any given point, and recovery from hypomania was more than 4 times as likely. The study defined recovery strictly: at least 8 consecutive weeks with no symptoms, or only 1 or 2 mild symptoms and no impairment in daily function.

One caution. This was an observational study. Treatment "is not assigned by design and not controlled by anyone connected with the study." Participants got whatever care they chose, using the medications available from the late 1970s onward. The figures describe episodes under real-world treatment of that era, not untreated illness and not modern care.

Why does bipolar depression last longer than mania?

The same research group looked at the question a different way. Instead of counting episodes, Judd and colleagues tracked mood week by week. Their 2002 paper in Archives of General Psychiatry followed 146 people with bipolar I for a mean of 12.8 years.

Participants were symptomatic in 47.3% of all weeks. Depressive symptoms accounted for 31.9% of follow-up weeks. Manic or hypomanic symptoms accounted for 8.9%, and cycling or mixed symptoms for 5.9%. Depression took up more than 3 times as many weeks as mania and hypomania combined.

The study also found that low-grade symptoms outlasted full episodes. Subsyndromal, minor depressive, and hypomanic symptoms together filled 29.9% of weeks. Full-criteria major depression and mania filled 11.2%. The long tail of a bipolar episode is often a stretch of milder depression, not the dramatic phase that got noticed.

Mania tends to be brief for practical reasons. It is visible, it disrupts sleep and behavior, and it often leads to urgent treatment or hospital care. Depression is quieter. People and families may wait longer before seeking help, and the episode itself resolves more slowly. If your depressive phases are the part that keeps you stuck, depression.md covers the treatment side in more depth.

What about mixed and rapid-cycling episodes?

Not every episode is a clean stretch of one mood. In the Solomon study, some participants moved between mania or hypomania and depression without 8 weeks of recovery in between. Researchers counted these as single "cycling" episodes, and they were the longest by far.

Major cycling episodes: median 42.0 weeks
Mixed major cycling episodes: median 61.0 weeks

Recovery odds were 56% lower for major cycling episodes and 67% lower for mixed major cycling episodes than for major depression. The authors note that cycling episodes lasted 3 to 20 times longer than pure mood states. Our guide to mixed episodes explains what this looks like from the inside.

Rapid cycling is a separate concept. NIMH defines it as "four or more episodes of mania or depression within a year." The individual episodes in a rapid-cycling pattern are often shorter, but there is little well time between them. The total burden of illness over a year can be higher, not lower. See rapid cycling bipolar disorder for how doctors approach it.

What makes an episode last longer?

The Solomon study identified two factors that predicted slower recovery, independent of episode type.

Severe onset. Episodes that began with psychosis or severe loss of function in the first week were less likely to resolve at any given point. A hard start tends to mean a long course.

Cumulative morbidity. People who had already spent more time ill during the study recovered more slowly from later episodes. Past illness burden predicts future episode length.

Episode type matters as much as anything. As the medians above show, a depressive episode is expected to run about twice as long as a manic one, and a cycling episode several times longer still.

The study could not measure how much treatment shortened episodes, because treatment was not controlled. What is clear from the diagnostic criteria is that a manic episode severe enough to need hospital care counts as mania at any duration. Getting care early, before that point, is the practical goal. Knowing your own episode triggers helps you and your care team act sooner.

How much time do people spend well between episodes?

More than half. In the Judd weekly-status study, participants were asymptomatic in 52.7% of weeks across nearly 13 years of follow-up. Bipolar I is a recurring illness, but it is not a constant one.

The recovery definition used in these studies matters here. An episode was counted as over only after 8 consecutive weeks of wellness or near-wellness. That means a person can feel better for a month and still be inside the same episode by research standards. It also means the 13-week median is a conservative figure. It includes the full stretch from first symptoms to sustained recovery.

For people with bipolar II or cyclothymia, the shape is different. MedlinePlus notes that cyclothymic symptoms "usually last for at least two years in adults and for one year in children and teenagers." The individual swings are milder, but the pattern is long.

The bottom line

A bipolar episode has to last at least 7 days (mania), 4 days (hypomania), or 2 weeks (depression) to meet diagnostic criteria. In practice, the median episode in the longest US study lasted 13 weeks, and about 3 in 4 people recovered within a year. Depression lasts about twice as long as mania. Cycling and mixed episodes last many times longer. Severe onset and a heavier past illness burden both slow recovery.

If an episode has run longer than these figures, that is worth raising with your prescriber. Long or cycling episodes are the ones where a change in treatment plan is most likely to help.

Last updated: September 2026. This article is for informational purposes only and does not constitute medical advice. Talk with your psychiatrist or care team about your own episode pattern and treatment plan. If you or someone you know is in crisis, call or text 988 in the US.