A variety of outcome measures assessing antidepressive therapy are available. However, in randomized clinical trials, the Hamilton Rating Scale for Depression (HAM-D) is often the primary outcome measure. Results from factor analysis and Rasch item analysis indicate that the HAM-D is heterogeneous and that the sum of items scores may not be an adequate measure of the severity of depression. A Melancholia Scale of 11 items has been suggested as a more valid measure of the core symptoms of affective syndrome. Other global outcome measures, focusing on health-related quality of life issues and on social functioning as well as macro-economic analyses are also used in depression. Applying stringent and well-documented outcome measures in randomized clinical trials of antidepressants may give the clinician a better indication of the most appropriate drug for treatment of the individual patient.