Bipolar disorder is a pathological disturbance of mood, characterized by waxing and waning manic, depressive and, sometimes distinctly mixed states. A diagnosis of bipolar disorder can only be made with certainty when the manic syndrome declares itself. Most individuals who are diagnosed with this disorder will experience both poles of the illness recurrently, but depressive episodes are the commonest cause of morbidity and, indeed, of death by suicide. Twin, adoption and epidemiological studies suggest a strongly genetic aetiology. It is a genetically and phenotypically complex disorder. Thus, the genes contributing are likely to be numerous and of small effect. Individuals with bipolar disorder also display deficits on a range of neuropsychological tasks in both the acute and euthymic phases of illness and correlations between number of affective episodes experienced and task performance are commonly reported. Current self-report and observer-rated scales are optimized for unipolar depression and hence limited in their ability to accurately assess bipolar depression. The development of a specific depression rating scale will improve the assessment of bipolar depression in both research and clinical settings. It will improve the development of better treatments and interventions. Guidelines support the use of antidepressants for bipolar depression. With regard to the adverse effects of antidepressants for bipolar depression, double-blind, placebo-controlled data suggest that antidepressant monotherapy or the addition of a tricyclic antidepressant may worsen the course of bipolar disorder. Importantly, adjunctive psychotherapies add significantly (both statistically and clinically) to the efficacy of pharmacological treatment regimens. The successful management of bipolar disorder clearly demands improved recognition of bipolar disorder and effective long-term treatment for bipolar depression as well as mania.