A new paradigm of the bipolar spectrum is shaping up in the research literature and in clinical practice. It represents a partial return to the Kraepelinian broad concept of manic depression which included many recurrent depressions. Although bipolar I and bipolar II are now part of the official nomenclature of Diagnostic and Statistical Manual for Mental Disorders, Fourth Edition (DSM-IV), the breadth of bipolarity is not represented in this manual. Old and new evidence indicate that the most common form of bipolar II is characterized by hypomanias of shorter duration than the arbitrary threshold of four days, and that cyclothymic depressions represent a prevalent variant of the bipolar II pattern. Furthermore, evidence is now compelling that hypomania in association with antidepressant treatments requires familial bipolar diathesis for bipolar disorder (bipolar III). There also exist clinical depressions superimposed on hyperthymic temperament (bipolar IV), referring to individuals with subthreshold hypomanic traits rather than episodes. Given emerging data for a population prevalence of at least 5% for a broadly conceived bipolar spectrum-and 6% for the cyclothymic temperament-the author submits that one of ten individuals in the community either has bipolar disorder, or is at risk for it. These are probably conservative estimates, because the range of clinical phenotypes and that of temperaments at risk are expanding. A provocative development is the emergence of extensive clinical and research evidence for the comorbidity of panic, social phobia, and related anxiety states with bipolar, and especially bipolar II, disorder. In addition, prevalent mixed states beyond dysphoric mania have been described, consisting of hypomanic intrusions into major depressive states.The net effect of the broadened boundaries of bipolarity is encroachment into the terrain of so-called "unipolar" anxious-depressions and of axis II cluster B. This is an evolving reformulation of the subclassification of affective disorders reviewed in this article, validated on the basis of phenomenology, epidemiology, course, family history, twin studies, and molecular genetics. These considerations have major implications for clinical practice, methodology of genetic investigations, pharmaceutical trials of putative bipolar agents in affective disorders, and public health.