Bipolar 1 and bipolar 2, and what a questionnaire can tell you

What separates the two types

If you are not safe right now, use the crisis numbers above before you read any further. Nothing below is urgent by comparison.

Both types describe the same territory: stretches of unusually high or irritable mood and energy set against stretches of low mood and energy. What separates them is how high the high stretch went, how long it ran, and what it did to a life while it was running.

The National Institute of Mental Health puts the first type in one sentence: bipolar I disorder is defined by manic episodes that last at least 7 days, most of the day and nearly every day, or when manic symptoms are so severe that hospital care is needed. The same page adds that separate depressive episodes usually occur as well, typically lasting at least 2 weeks.

The second type is defined against the first. Bipolar II disorder is defined by a pattern of depressive and hypomanic episodes, but the hypomanic episodes are less severe than the manic episodes in bipolar I disorder.

Read those two descriptions beside each other and the separating facts are visible. A duration of at least seven days. A severity threshold, which is whether the state was severe enough that hospital care was needed. And a comparison, which is whether the high stretches were less severe than that. All three are facts about stretches of time that have already happened, and two of them are judgements about how far something went rather than whether it happened at all.

One thing neither definition contains is a ranking of how hard the two are to live with. The number names which kind of high stretch is in the history. That is all it names.

Why the separation is a judgement about a history

Once you know what the two definitions turn on, it becomes clear why no form settles it. The same NIMH page says that bipolar disorder is diagnosed based on the severity, length and frequency of an individual's symptoms and experiences over their lifetime, and their family history. Lifetime is the operative word. A questionnaire answered this afternoon samples this afternoon's memory of a lifetime.

The route is a referral rather than a result. NIMH describes a health care provider completing a physical exam, ordering medical testing to rule out other illnesses, and referring the person for an evaluation by a mental health professional. In the UK the NHS describes the same shape: if a GP thinks you may have bipolar disorder or another mental health condition, they will refer you to a mental health specialist, and that specialist will ask about things like your moods, behaviour, health and family history.

The NHS is also plain about the difficulty, saying that bipolar disorder can take time to diagnose because it affects everyone differently and the symptoms are similar to other mental health conditions.

There is a further reason the high stretches in particular are hard to establish from self-report, and a review of bipolar screening tools names it directly: patients' hesitation to report hypomanic or manic symptoms. The same review notes that the disorder often begins with a depressive episode, leading to misdiagnosis as other conditions, and that patients typically take an average of 8 to 10 years to obtain an accurate diagnosis.

Put those together and the practical problem is specific. The stretch that decides the type is the one a person is least placed to describe, because from inside it can read as a good month rather than as an episode. It is also the one other people remember most clearly. That is why an account from a partner, a parent, a sibling or an old friend is worth so much in this particular conversation, and why a browser tab with one respondent in it cannot stand in for one.

The published screeners decline the separation, and say so

This is the part most pages that offer a type 1 or type 2 quiz leave out. The instruments the field actually uses do not claim to make the separation, and their own literature is explicit about it.

A 2025 review of screening tools for bipolar disorders describes the Hypomania Checklist as particularly effective for distinguishing bipolar disorder from unipolar depression, although it cannot differentiate between bipolar disorder type I and bipolar disorder type II. That is the best-known hypomania instrument in the field saying it does not do the one thing a typed quiz would have to do.

The same review says the Mood Disorder Questionnaire often exhibits lower sensitivity for detecting bipolar disorder type II, particularly in individuals with mild or subthreshold hypomanic symptoms. Read that slowly: the instrument is weakest on exactly the presentation that defines the second type.

The review is equally clear about what a screening result is for. As a screening tool, it aims to identify potential cases, which clinicians should further investigate to confirm or rule out the positive screening result, and positive results should always be followed by a thorough clinical evaluation. It goes further still on where screening belongs, noting that routine screening for depression is recommended at the primary care level, while screening for bipolar disorder is not.

So the position is not that these are good instruments we happen not to license. It is that the instruments themselves are designed to hand a question to a clinician, and that the question they hand over is whether bipolar disorder is worth investigating at all, not which of two types a person has.

What a positive screen is worth in a general audience

There is a second problem, and it is arithmetic rather than wording. A screening instrument behaves differently depending on how common the thing is in the group being screened, and a website's audience is not a psychiatric clinic's waiting room.

A large community study of the Mood Disorder Questionnaire measured this directly. For bipolar disorder, sensitivity was moderate at 42.9% and specificity was high at 96.2%. Those two figures look respectable, and the third one is the one that matters: the positive predictive value was very low, at 0.033.

In plain terms, in that community sample, roughly three people in every hundred who screened positive turned out to have bipolar disorder. The other ninety-seven did not. The study also reports that a large proportion of the people who screened positive were found to have anxiety disorders, major depression, or substance use disorders. Its own summary sentence is blunter than anything this site would write: it confirms the uselessness of the questionnaire as a screener for bipolar disorder in that setting.

Note that the errors run both ways, and the second one is quieter. A sensitivity of 42.9% means the same instrument missed more than half of the people who did have it. A reassuring negative from a short questionnaire is not evidence of anything, and treating it as evidence is how a real conversation gets postponed.

This site does not reproduce the Mood Disorder Questionnaire or the Hypomania Checklist. Neither is ours to reproduce, and on the evidence above neither would type anyone if it were.

Why there is no type 1 test and no type 2 test here

The demand is real and the search is reasonable. Someone who has read the two definitions wants to know which one describes them, and a page offering to answer that is easy to build. Here is why this site will not build two of them.

The separation rests on how long a stretch ran and how severe it got, years after the fact. A rating scale asks how strongly a statement describes you. Those are different kinds of question, and no amount of careful wording turns the second into the first.

A label of this kind is not a neutral piece of information. It changes how a person reads their own history, how the people around them read it, and what they expect from an appointment. Handing that to a stranger on the strength of five minutes and no history is not a small mistake made carefully, it is the wrong shape of product.

And a typed result would be doubly wrong, because it would compound a screening judgement this site cannot make with a typing judgement the published instruments decline to make. Two errors stacked on each other do not average out.

What can honestly be published is this guide, and a self-check that describes patterns without naming a condition. That is the whole of the offer.

What the self-check on this site does instead

MyTestAtlas has one mood route, the Mood and Energy Patterns Self-Check. It asks about eighteen everyday situations from the last year or two and reports four patterns separately: high-energy stretches, low-energy stretches, shifts and cycles, and costs and after-effects. There is no total, no cut-off, no type and no verdict.

The fourth pattern is the one doing the honest work. A stretch of unusual drive with nothing following it is a good fortnight. The same stretch with a purchase, a resignation or a fortnight of apologies behind it is the thing worth describing to a clinician, and keeping the two apart on the page is what a total would destroy.

The value of it is language rather than an answer. People often arrive at an appointment able to say only that their mood is all over the place. Arriving instead with four named patterns, and two or three dated examples behind each, turns that into something a clinician can work with in a short appointment.

What it cannot do is separate these patterns from the things that produce the same answers. The route says so on its own page: grief, thyroid trouble, alcohol or other substances, medication effects, broken sleep, ADHD, anxiety and an ordinary reaction to a hard year all read alike on a questionnaire, and only a clinician working from a history over years can weigh them.

And if you are not safe, the crisis numbers at the top of this page matter more than anything else on it.