Neurobehavioral sequelae of strokes can limit a patient’s ability to describe or express emotion, can cause him to give “yes” answers to the clinician who expects them, or can directly cause apathy or crying spells. Also, anosognosia for depressive signs can cause the patient to deny depressive signs that are objectively observable. These diagnostic confounders have not been adequately assessed in previous research on poststroke depression; thus many studies are of doubtful validity, as shown by studies of the dexamethasone suppression test for melancholia in stroke patients. Future studies on depression after stroke must prospectively rule out fluent aphasia, motor aprosody, and amnesia before relying on diagnostic information from the psychiatric interview, and the interview should always be supplemented by direct observation of vegetative signs and other behavior. With this extended information, major- depression can and should be diagnosed using accepted symptom and duration criteria.