A case-control study was conducted on 303 male bladder cancer patients and controls. General population controls were chosen from 15 areas in Gunma Prefecture and were matched by age (+/- l y.o.) to the subjects. Age-adjusted and smoking-adjusted odds ratio (O.R.) and a 95% confidence interval (C.I.) were calculated for each item. Risk factors for bladder cancer in men were investigated. The O.R. tended to be significantly higher for those who had history of smoking, who smoked more per day, who had smoked longer, whose Brinkman index was higher, who began smoking younger and who inhaled deeper than it was for non-smokers. O.R.s of having a past history or complication of cystitis (age-adjusted) and benign prostatic hypertrophy (age- and smoking-adjusted) were significantly higher, but the difference was supposed to be caused by bias. There was a significantly lower age- and smoking-adjusted O.R. for bladder cancer in men who engaged in sales, whose blood type was O, who drank milk frequently, who ate grains frequently, who age vegetables frequently and who had a past history or complication of hypertension. The number of cases and controls with first degree family members who developed cancer respectively supposed to be highly related to smoking, were as follows; 16 and 8 for lung cancer, 3 and 0 for larynx cancer and 6 and 3 for bladder cancer. The following characteristics failed to show any significant difference between subjects with bladder cancer and the control group; height and weight now and 20 years ago, jobs which deal with dye, academic career, marriage, number of children, alcohol drinking and the use of hair dye or analgesics.