In spite of the availability of antidepressant medication for several decades, it has not been shown that such medication lowers the risk for suicide in depressed patients. This report explores this apparent paradox by means of pharmacoepidemiological methods. Data on the prevalence of depression in the population and among suicides as well as data on the prevalence of antidepressant medication in depressed suicides were obtained from a review of the literature. Data on the prevalence of antidepressant medication in the population in 1990-1991 were obtained from the statistics of the Swedish National Corporation of Pharmacies. It was found that only one in five depressed individuals with major depression were treated with antidepressants in Sweden. The calculated risk for suicide among depressed patients who were treated with antidepressants was 141 per 100,000 person years and, among the untreated, 259 per 100,000 person years (i.e., 1.8 times higher among the untreated). This supports the hypothesis that antidepressant medication decreases the risk for suicide in depressed patients. The reason this has not been obvious in the general suicide statistics seems to be that so few depressed people are treated with antidepressants. Effective suicide prevention strategies should include intensive efforts to recognize and treat more depressed people.