The diagnostic validity of atypical depression is based on its superior response to monoamine oxidase inhibitors compared to tricyclic antidepressants, and on latent class analysis. The studies on atypical depression have often not included bipolar patients. The aim of the present study was to find the prevalence of bipolar II disorder among DSM-IV atypical depression outpatients. Bipolar II and unipolar atypical depressions were also compared to find if they were variants of the same disorder or if instead they were different disorders. One hundred and forty consecutive unipolar and bipolar II outpatients, presenting for treatment of an atypical major depressive episode, were interviewed with the Structured Clinical Interview for DSM-IV, the Montgomery Asberg Depression Rating Scale (MADRS), and the Global Assessment of Functioning Scale. The prevalence of bipolar II disorder was 64.2%. The age at baseline and onset were significantly lower in bipolar II versus unipolar patients. All the other variables (MADRS items, duration of illness, severity, gender, psychosis, comorbidity, chronicity, recurrences) were not significantly different. The prevalence of bipolar II disorder among atypical depressed outpatients was higher than previously reported.