Patients with abdominal sepsis were enrolled in a clinical trial of aztreonam vs. tobramycin. All were given clindamycin concomitantly. The pharmacokinetics of aztreonam in 21 patients randomly assigned to receive treatment with aztreonam are reported. The mean age of these patients was 68 years; most had underlying disorders such as malnutrition and cardiac or pulmonary disease. Creatinine clearance (Clcr) ranged from 11.2 to 133.1 ml/min. The usual dose of aztreonam was 2.0 g every 8-12 hr. A single pharmacokinetic study was performed over one dosing interval after steady-state conditions were achieved. In approximately one-half of the patients, peritoneal fluid was collected during the interval between doses. Penetration of aztreonam, as expressed as the ratio of concentration in the peritoneal fluid to that in serum, was higher for aztreonam (0.95:1) than for tobramycin (0.46:1). The ratio of the concentration in peritoneal fluid to the minimum inhibitory concentration (MIC) of the infecting bacteria was also higher for aztreonam. Serum pharmacokinetic data were analyzed by both two-compartment and moment analysis. For both the steady-state volume of distribution (Vdss) and total body clearance (TBC), the values determined by both methods were highly correlated (r = .96, .99, respectively). Average values for Vdss and TBC were 0.28 liters/kg and 80 ml/min. TBC for aztreonam correlated strongly with CLcr and was described by the regression equation TBC = 1.1 (Clcr) + 1.6, r = .87, P less than .01.