To study once-daily antimicrobial prophylaxis of urinary tract infections, we gave trimethoprim-sulfamethoxazole (40 mg/200 mg), trimethoprim (100 mg), nitrofurantoin macrocrystals (100 mg), or placebo to 60 women for 6 months. During prophylaxis, infections per patient year were comparable in the groups receiving trimethoprim (0.0), nitrofurantoin (0.14), or trimethoprim-sulfamethoxazole (0.15) and occurred less frequently than in patients receiving placebo (2.8; P less than 0.001, placebo versus each drug regimen). The effectiveness of prophylaxis was limited to the 6 months that antimicrobials were given, and infections were more likely to develop after prophylaxis in women who had three or more infections in the year before prophylaxis (P less than 0.005). Further, women whose preprophylaxis infection was positive for antibody-coated bacteria were more likely to have same-strain relapse when infections recurred (P = 0.001). Emergence of trimethoprim-resistant Escherichia coli was rare, but non-E. coli infections occurred more often after prophylaxis (P less than 0.05). Prophylaxis with these drugs is effective, well tolerated, and did not produce emergence of resistant E. coli but may predispose to non-E. coli urinary tract infections after its discontinuation.