Abstract
1. Methodological reasons are discussed which make placebo‐controlled trials on antidepressant drugs necessary. 2. The ethical problem of treating a sick patient with an inactive compound are shown to have an impact on the methodology of such trials. 3. Three double‐blind placebo‐ controlled trials with nomifensine are reviewed; one in neurotic depression, one in psychotic depression and one in geriatric patients showing depressive symptoms. 4. In all three trials the effect of nomifensine has been shown to be superior to that of placebo.