The relationship of age, medical history, personal habits, and urologic symptoms to the incidence of surgically treated benign prostatic hyperplasia (BPH) was studied in a cohort of 16,219 men, aged forty years and over, who received multiphasic health checkups (MHCs) during 1971 and 1972 in Oakland or San Francisco while members of the Northern California Kaiser Permanente Medical Care Program, a large prepaid health care program. Follow-up was carried out for surgically treated BPH from the date of the MHC to the date of the earliest of the following: surgery for BPH (n = 1,027); incidence of prostate cancer (n = 329), bladder cancer (n = 119), or both (n = 10); other prostate surgery (n = 5); death (n = 2,525); membership termination (n = 4,235); or December 31, 1987 (n = 7,969). The mean length of follow-up was twelve years. In multivariate analysis utilizing the Cox proportional hazards model, the following characteristics were positively associated (p less than 0.05) with risk of surgically treated BPH: age, low body mass index, nonsmoking (vs. current smoking), urine pH greater than 5, history of kidney x-ray and of tuberculosis, and each of five urologic symptoms (dysuria, loss of bladder control, trouble starting urination, nocturia, slow urine stream). The risk of BPH associated with obstructive urologic symptoms decreased markedly with age. Some of these findings are consistent with those from other studies (age, nonsmoking), while others (high urine pH, history of tuberculosis) are new and should be examined in other study populations.