Bereavement is a major risk factor for physical illness, grief, depression, and anxiety. In contrast to recent tendencies in the psychiatric literature to equate grief and depression, we propose that a careful discrimination between the two must be made for diagnostic, therapeutic, and investigative purposes. We report the results of a longitudinal study of a frequent but neglected event, miscarriage early in pregnancy, to make this point. Clinical criteria for differentiating grief and depressive reactions were developed based on phenomenological criteria and theoretical considerations. We hypothesized that the detrimental psychological and physical consequences occur only when the miscarriage was not mourned and resulted in a depressive reaction, but not in a grief reaction. In a controlled, representative study, 125 consecutive women were assessed shortly after their miscarriage (before the 20th week of gestation) and 6 months (N = 94) and 12 months (N = 90) later. Assessments included standardized questionnaires for life events, depression, physical complaints, anxiety, and a specific, multidimensional grief scale (Munich Grief Scale) that we had developed previously. Immediately after the miscarriage, the average anxiety and depression scores were elevated when compared with 80 pregnant and 125 age-matched community controls. Twenty percent of the patients who had miscarried showed a grief reaction, 12% showed a depressive reaction, and 20% responded with a combined depressive and grief reaction. The remaining women (48%) reported no changes in their emotional reactions. As predicted, longer-lasting psychological, social, and health status changes followed the initial depressive, but not the grief reactions. Depressive reactions were predicted by a history of previous depression, a lack of social resources, and an ambivalent attitude to the lost fetus. The grief measures were reliable and made it possible to discriminate between grief and depression.