Bipolar disorder constitutes a challenge for clinicians in everyday clinical practice. Our knowledge concerning this clinical entity is incomplete and contemporary classification systems are unable to reflect the complexity of this disorder.. The concept of temperament which for the first time was described during antiquity constitutes a reasonable vehicle to synthesize our knowledge on how the human body works and what determines human behavior. Although it originally included philosophical and sociocultural approaches, today the biomedical model is dominant. It is possible that specific temperaments might constitute vulnerability factors, determine the clinical picture or constitute illness course modifiers and even act as a bridge between genes and clinical manifestations, thus giving birth to the concept of the bipolar spectrum with major implications for all aspects of mental health research and providing of care. More specifically it has been reported that the hyperthymic and the depressive temperaments are related to the more 'classic' bipolar disorder, while cyclothymic, anxious and irritable temperaments are related to more complex manifestations and might predict poor response to treatment, violent or suicidal behavior and high comorbidity. It seems reasonable to assume that the incorporating of the concept of temperament and the bipolar spectrum in the standard training of psychiatric residents might result in an improvement of everyday clinical practice.3