The efficacy of oral ciprofloxacin was compared with that of parenteral aminoglycoside for therapy of complicated urinary tract infection in a prospective randomized trial. The setting was a chronic-care Veterans Administration facility in which long-term bladder catheterization and resistant bacteria were common. Sixty-five consecutive patients were stratified for presence and type of bladder catheter, the presence of Providencia and Pseudomonas aeruginosa organisms versus other urinary pathogens, and renal dysfunction. A pharmacokinetic study of ciprofloxacin concentrations in serum and urine was performed with selected patients. Urinary catheters were present in 94%. All patients were symptomatic, and 58.5% had fever. Ciprofloxacin, 500 mg every 12 h orally, was compared with parenteral aminoglycoside for 7 to 10 days. Clinical response, defined by resolution of symptoms and fever at 5 to 9 days posttherapy (short-term) and 28 to 30 days posttherapy (long-term), was essentially identical for both treatment groups. Bacteriologic response, defined by sterile urine cultures, showed that ciprofloxacin was significantly more efficacious (P = 0.0005) than aminoglycoside at 5 to 9 days posttherapy. However, by 28 to 30 days, the response rate was essentially identical. Emergence of resistance to the study antibiotic was seen in 62 and 70% of those who did not show a bacteriological response and were receiving ciprofloxacin and aminoglycosides, respectively. Concentrations of ciprofloxacin and aminoglycoside in the urine substantially exceeded the MIC for the urinary pathogens isolated, although these concentrations did not correlate with short-term bacteriologic response for either antibiotic. Ciprofloxacin was efficacious in complicated urinary tract infection compared with the current standard of parenteral aminoglycoside among catheterized patients with relatively resistant bacteria.