There is growing recognition that bipolar disorder (BPD) has a spectrum of expression that is substantially more common than the 1% BP-I prevalence traditionally found in population surveys. To estimate the prevalence, correlates, and treatment patterns of bipolar spectrum disorder in the US population. Direct interviews. Households in the continental United States. A nationally representative sample of 9282 English-speaking adults (aged >or=18 years). Version 3.0 of the World Health Organization's Composite International Diagnostic Interview, a fully structured lay-administered diagnostic interview, was used to assess DSM-IV lifetime and 12-month Axis I disorders. Subthreshold BPD was defined as recurrent hypomania without a major depressive episode or with fewer symptoms than required for threshold hypomania. Indicators of clinical severity included age at onset, chronicity, symptom severity, role impairment, comorbidity, and treatment. Lifetime (and 12-month) prevalence estimates are 1.0% (0.6%) for BP-I, 1.1% (0.8%) for BP-II, and 2.4% (1.4%) for subthreshold BPD. Most respondents with threshold and subthreshold BPD had lifetime comorbidity with other Axis I disorders, particularly anxiety disorders. Clinical severity and role impairment are greater for threshold than for subthreshold BPD and for BP-II than for BP-I episodes of major depression, but subthreshold cases still have moderate to severe clinical severity and role impairment. Although most people with BPD receive lifetime professional treatment for emotional problems, use of antimanic medication is uncommon, especially in general medical settings. This study presents the first prevalence estimates of the BPD spectrum in a probability sample of the United States. Subthreshold BPD is common, clinically significant, and underdetected in treatment settings. Inappropriate treatment of BPD is a serious problem in the US population. Explicit criteria are needed to define subthreshold BPD for future clinical and research purposes.