Angst (Reference Angst2007) provides more balanced views on the much publicised concerns about the underdiagnosis of bipolar disorder. Psychiatric diagnoses are not robust entities (Reference Baca–Garcia, Perez–Rodriguez and Basurte–VillamorBaca-Garcia et al, 2007) and most recent research in mood disorders has arisen from redefining and often rigidly applying the DSM criteria, which has proved a hindrance to research. The problem in mood disorder research lies in our failure to define the core features of mania/hypomania and bipolar depression. Surprisingly, hardly any advance has been made in our understanding of and our ability to accurately diagnose an active hypomanic/manic episode (excluding retrospective accounts), and we are guided by epidemiological studies and expert opinions rather than basing diagnosis on a new phenomenological understanding. Moreover, we rely on a range of self-report checklists. Unfortunately, there are few advocates for people with wrongly diagnosed bipolar disorder. It is like initiating antihypertensive treatment for suspected hypertension. Unless they have clinical consequences, temperament and vegetative lability, like blood pressure and heart rate, should not be considered pathological. The success of future research lies in a greater understanding of the phenomenology of episodes of depression and in bipolar disorder and the differences in biological depression that result from psychosocial factors.