Urinary incontinence after radical prostatectomy is a difficult postoperative problem and often is a major consideration in the selection of therapy for clinically localized disease. The occurrence of incontinence is unpredictable and the relationship of incontinence to operative technique is unclear. We compared urinary continence in 68 consecutive patients undergoing radical prostatectomy. In 34 patients nonnerve-sparing radical prostatectomy was performed and in 34 subsequent patients a nerve-sparing operation was done. Patient age, Gleason score and stage of the tumor, and operative time were not significantly different between the groups. In the nonnerve-sparing operated group there were 4 patients (12%) with total and 6 (18%) with stress incontinence requiring absorbent pads, compared to 0 and 2 (6%), respectively, in the nerve-sparing group. The postoperative functional urethral length in the nonnerve-sparing group was 1.9 +/- 0.6 cm. (standard deviation) and in the nerve-sparing group it was 2.3 +/- 0.5 cm., which was significantly different (p less than 0.05). The peak resting urethral pressure of the nonnerve-sparing group was 35.4 +/- 14.2 cm. water and in the nerve-sparing group it was 46.5 +/- 12.3 cm. water, which also was significantly different (p less than 0.05). The study indicates that preservation of the pelvic nerves during radical prostatectomy has a major role in the functional preservation of urinary continence.