Bipolar disorder or manic-depressive illness has been distinguished from schizophrenia since Kraeplin over 100 years ago, although a persistent group of psychiatrists, scientists, and accompanying facts seems to turn up in every generation to resurrect the concept of “a unitary psychosis” that may even include depression (3). The discovery of lithium as a simple ion effective in bipolar disorder and not in schizophrenia presaged an era in which the distinction between bipolar disorder and schizophrenia was a foundation of psychiatric treatment, and a prediction of lithium response a clinical tour de force to make any psychiatrist proud. The later discovery that some anticonvulsants, such as valproate and carbamazepine, are also effective in bipolar disorder did not undermine the essential concept. However, the rise of atypical antipsychotic use in bipolar disorder and the large quantity and quality of the evidence justifying its use over the last 10 years is having a major impact on clinical diagnostic thinking (4). This is not yet evident in DSM-5, but I predict that it will be evident in DSM-6. Most, if not all, of the new atypical antipsychotics are effective in mania and in the prophylaxis of new episodes in bipolar disorder. But what about bipolar depression? Bipolar depression is a major clinical problem, and the evidence that antidepressants can be useful seems less and less convincing (5). Like a brave knight on a white horse, atypical antipsychotics have rushed in to save us, the psychiatric maidens. Quetiapine is clearly effective in bipolar depression, and its evidence base is backed up by clinical satisfaction. However, weight gain, hyperlipidemia, and glucose impairment are striking, although not as striking as with olanzapine, another atypical antipsychotic clearly shown to be useful not only in bipolar mania and in the prophylaxis of bipolar episodes but also in the treatment of bipolar depression with or without an adjunctive mood stabilizer.