To determine the clinical and therapeutic relevance of longitudinally predominant polarity for bipolar disorders long-term outcome.Two hundred twenty-four patients (n=224) were enrolled for the study in the Bipolar Disorders Program of Barcelona, which provides integrated care for difficult-to-treat bipolar patients derived from all over Spain, but also provides clinical care to all bipolar patients coming from a specific catchment area (Eixample Esquerre) in Barcelona. Data collection regarding predominant polarity started on October 1994 and lasted for the following ten years. Patients were divided according to the predominance of depressive or manic/hypomanic episodes. The two groups were compared regarding clinical and sociodemographic variables.135 patients (60.3%) were classified as Depressive Polarity, whilst 89 (39.7%) were considered as Manic Polarity. Manic Polarity was more prevalent amongst bipolar I patients than bipolar II. Depressive Polarity was strongly associated with depressive onset of bipolar disorder. Lifetime history of attempted suicide was strongly associated with Depressive Polarity, who also had a higher mean number of suicide attempts. As for therapeutic issues, acute and maintenance use of atypical antipsychotics and conventional neuroleptics were more common amongst Manic Polarity whilst antidepressants and lamotrigine use was highly prevalent amongst Depressive Polarity.Prevention of depression is crucial for the maintenance treatment of bipolar II patients, whilst prevention of mania and depression would be equally important in the case of bipolar I patients. Predominant polarity is a valid prognostic parameter with therapeutic implications.