There is little rationale to support the currently accepted 10 to 14-day treatment period for urinary tract infection. To assess the cure rate in patients given 3 days compared to the standard 10-day period of antimicrobial therapy a prospective trial was designed to test the over-all effectiveness of each treatment. Sixty patients were randomized to either penicillin-G or trimethoprim-sulfameth-oxazole for either 3 or 10 days. Urine and external vaginal cultures were done before therapy was instituted, at day 3 while on therapy and 7 days after the completion of treatment. The over-all cure rate in the short-term treatment group was 86 per cent and in the long-term treatment group it was 88 per cent. Patients receiving trimethoprim-sulfamethoxazole had a much better response to clearance of the pathogenic bacteria from the external vagina but this did not correlate with clearance of the bladder bacteriuria. Bacterial sensitivities on the external vaginal cultures suggest that in some patients 10 days of therapy actually may aid in the development of bacterial resistance noted in subsequent bladder infections. Finally, a cost-benefit analysis revealed that the use of a 3-day regimen as standard treatment for urinary tract infections would result in a savings to our patients conservatively estimated at $62,000,000 yearly.