THIS PAPER reports longitudinal steroid data on nine patients who subsequently committed suicide or who made a serious attempt and compares the significance of the elevations of these levels with 134 other individuals studied in our own laboratory and reviewed from the psychiatric literature. Behavioral and biochemical studies of severely depressed inpatients suggested the hypothesis that an association might exist between high urinary 17-hydroxycorticosteroids (17-OHCS), which were utilized as an index of psychological distress, and increased suicidal risk. This hypothesis was initially based on a retrospective study1of three depressed patients who committed suicide and a comparison group of 33 depressed impatients. These studies occurred during the course of five years of study which involved contact with 250 depressed patients, 36 of whom were treated in an inpatient setting. In the three years since this initial report, longitudinal steroid data have