A comparative double-blind study of the clinical efficacy of cefoperazone against chronic, complicated urinary tract infections was performed using carbenicillin as a control. Cefoperazone was administered at 2 gm/day and carbenicillin at 4 gm/day, in two divided doses, for five consecutive days, and the clinical efficacy was determined. A total of 283 patients were treated, of which 45 were subsequently excluded as unassessable ans six were dropped owing to side effects, leaving 232 evaluable patients, 116 on each drug. There were no significant differences between the two drug groups with respect to patients' sex, age, body weight, and catheterization. The overall efficacy rates were: cefoperazone, excellent in 24 and moderate in 45 cases, for an overall clinical effectiveness rate of 59%; carbenicillin, excellent in 14 and moderate in 21 cases, for an overall effectiveness rate of 30%. With respect to bacteriological activity, 68.2% of 154 strains treated with cefoperazone were eradicated, compared with 50.3% of 169 strains treated with carbenicillin. The corresponding eradication rates were 84% and 40% for Escherichia coli, 50% and 16% for Serratia sp, and 54% and 55% for Pseudomonas sp, respectively. Cefoperazone caused side effects in only four cases, and carbenicillin in three, all of which were mild in nature, with no significant toleration differences between the two groups. A total of 11 abnormal laboratory test results were obtained in patients treated with cefoperazone; nine such results were recorded for carbenicillin-treated individuals. None were of clinical significance.