Seventy-seven adults with urinary tract infections were randomly assigned in double-blind fashion to receive cefepime or ceftazidime 500 mg intravenously every 12 hours. Forty-eight patients (26 in the cefepime group and 22 in the ceftazidime group) were assessable bacteriologically. Groups were similar in mean age (cefepime, 60.6 years; ceftazidime, 63.4 years), sex distribution (cefepime, 77% male; ceftazidime, 59% male), frequency of complicated urinary tract infection (cefepime, 69%; ceftazidime, 77%), frequency of pyelonephritis (cefepime, 15%; ceftazidime, 18%), and days of therapy (cefepime, 10.3 days; ceftazidime, 10.7 days). Pretherapy isolates were Escherichia coli (cefepime, 16 patients; ceftazidime, 14 patients), Pseudomonas aeruginosa (cefepime, 3; ceftazidime, 5), Klebsiella species (cefepime, 4; ceftazidime, 3), Enterobacter aerogenes (cefepime, 3), Citrobacter freundii (cefepime, 1), and Morganella morganii (ceftazidime, 1). Using the criterion of < 10(4) colony-forming units/mL, 22 (85%) of 26 cefepime patients and 20 (91%) of 22 ceftazidime patients were cured 1 week posttherapy (P = 0.67). Early failures were due to E coli (cefepime, 3; ceftazidime, 2) and P aeruginosa (cefepime, 1). At 4 to 6 weeks posttherapy, 12 (71%) of 17 cefepime patients and 12 (63%) of 19 ceftazidime patients remained cured (P = 0.73). Adverse events were mild to moderate, reversible, and occurred in 6 (15%) of 39 cefepime patients and 4 (11%) of 38 ceftazidime patients (P = 0.74). Cefepime appears to be as safe and effective as ceftazidime in the treatment of adults with urinary tract infections.