Since 1981 we have been studying prostate cancer (Pca) by mass screening (MS) in the Gunma Prefecture, Japan. From 1981 to 1990, 9,067 subjects (total 15,451) were examined in connection with this project and 121 subjects were diagnosed as having Pca. The presence of Pca in 87 cases was confirmed at their initial MS (initial group); 34 cases were confirmed either the year after or several years after the first MS (repeat group). To evaluate the significance of MS for Pca, in this group of patients, the effects of ‘lead time bias’, ‘self-selection bias’ and ‘length bias’ on the survival rate were compared to Pca patients detected in the hospitals of the Gunma Prefecture during the same period. The survival curves of MS cases for each stage of the disease were better than those of the controls. Only in stage D was there a significant difference between the two groups. However, this MS curve decreased from the 4th year in the same manner as the control curve from the start. The clinical characteristics (age, pathological differentiation, prostatic acid phosphatase, gait disturbance, erythrocyte sedimentation rate, chronic disease, and pain in the patients with stage D disease) were compared between both groups. All characteristics studied, except age, in the MS group were clinically more favorable compared to those of the controls for each stage or in total. Moreover, the relative survival rate of subjects examined by MS was greater than 1.0. No significant difference was found in stage distribution between the initial group and the repeat group. During the follow-up study, 15 Pca patients were found who had a history of MS examination but were diagnosed in standard hospitals. Only 2 of these 15 patients were diagnosed with Pea within 1 year following their last MS examination, and they had stage A disease. We conclude that MS should be performed annually and recommend that people who want an early diagnosis should receive MS once a year.