Background: A recent series of studies has questioned the current categorical split of mood disorders into bipolar and depressive disorders. Mixed states, especially mixed depression (i.e., depression plus co-occurring, noneuphoric, hypomanic symptoms) might support a continuity between bipolar II (BP-II) depression and major depressive disorder (MDD). The aim of the study was to assess the distribution of intradepressive hypomanic symptoms rating between BP-II and MDD depressions. A bi-modal distribution would support a categorical distinction, and no bimodality would support continuity. Methods: Consecutive 389 BP-II and 261 MDD major depressive episode (MDE) outpatients were interviewed (off psychoactive drugs) with the Structured Clinical Interview for DSM-IV, the Hypomania Interview Guide (HIG, to assess intradepressive hypomanic symptoms), and the Family History Screen, by a mood specialist psychiatrist in a private practice. Mixed depression was defined as MDE plus 3 or more intradepressive, noneuphoric hypomanic symptoms, a definition validated by Akiskal and Benazzi. The distribution of intradepressive hypomanic symptoms rating was studied by Kernel density estimate and by histogram. Results: BP-II depression, versus MDD depression, had significantly lower age at onset, was significantly more likely to be atypical and mixed, had more depression recurrences, and a higher bipolar family history loading. BP-II depression, versus MDD depression, had significantly more irritability, racing/crowded thoughts, distractibility, psychomotor agitation, talkativeness, increased goal-directed activity, and excessive risky activities. HIG scores were significantly higher in BP-II. The distribution of intradepressive hypomanic symptoms rating showed no bi-modality in the entire depression sample. Conclusions: Interpretation of study findings relies on the method used to define a categorical disorder. By using classic diagnostic validators (such as family history and age at onset), BP-H and MDD depressions would seem to be distinct disorders. Instead, by using the 'bi-modality' approach, a continuity would seem to be supported. Which of these methods for classification is the best has yet to be shown. (c) 2006 Elsevier Inc. All rights reserved.