Patients with long-term indwelling urethral catheters are subject to acute and long-term complications of bacteriuria. To evaluate the common practice of short-course antibiotic therapy in such patients, we performed a randomized controlled trial of ten-day courses of cephalexin monohydrate repeated whenever a susceptible bacteriuria was present. We observed 17 cephalexin group patients for 545 patient-weeks (160 cephalexin courses) and 18 control group patients for 477 patient-weeks. Throughout the study, the groups were comparable in regard to incidence and prevalence of bacteriuria, number of bacterial strains per weekly urine specimen, incidence of febrile days, and incidence of obstructed catheters. In the cephalexin group, the frequency of fever during periods when antibiotics were being used was similar to that during periods when antibiotics were not being used. More cephalexin-resistant bacteria were isolated from cephalexin group patients. Routine treatment with cephalexin of asymptomatic long-term catheterized patients, even for susceptible organisms, does not seem to be warranted.