There are increasing levels of psychological distress among general practitioners (GPs). The purpose of this study is to evaluate the effectiveness of a mailed intervention to reduce distress among 'at-risk' GPs. A questionnaire was sent to 1356 GPs from eight Divisions of General Practice. Out of 819 (60%) who responded, 233 GPs were recruited with scores indicative of psychological distress. These GPs were randomized to intervention (n = 120) or control (n = 113). The intervention consisted of a simple letter feeding back and interpreting the psychological score together with a self-help sheet. During the study, an educational program was offered to GPs by Divisions of General Practice. The main outcome measure used was changes in psychological distress (General Health Questionnaire 12) score after 3 months. Significance was set at P < 0.05. Initial analysis of the data showed borderline significance (P = 0.05). However, analysis of the data post hoc excluding GPs who participated in the educational program showed a significant reduction in psychological distress (P = 0.03). It appears that there may have been a dilution of the intervention effect. Mailed interventions are a cost-effective way of reaching at-risk GPs and may contribute to a reduction in psychological morbidity.