OBJECTIVE To evaluate the pharmacokinetic profiles of aztreonam and imipenem in critically ill trauma patients with pneumonia. METHODS Trauma patients in intensive care units who were intubated within 3 days of hospital admission were eligible for the study. Patients with the clinical diagnosis of pneumonia were consecutively randomized to receive either aztreonam plus vancomycin or imipenem-cilastatin. Serial blood samples were taken and sputum was collected to determine aztreonam and imipenem concentrations after 2-3 days and 7-8 days of therapy. Pharmacokinetics of both agents were estimated and compared with estimates from healthy volunteers. RESULTS Twenty patients were enrolled in the study, 10 patients received imipenem-cilastatin, and 10 received aztreonam plus vancomycin. Steady-state volume of distribution (Vss) for aztreonam at 2-3 days and 7-8 days was significantly greater in patients than in historical controls, whereas the Vss for imipenem was greater at 2-3 days. The beta-half-life for aztreonam at both sampling periods was significantly greater in patients than in controls. No significant changes in pharmacokinetics occurred over time for either antibiotic. Sputum concentrations of aztreonam and imipenem were highly variable when sampled 2 hours after the infusion. CONCLUSIONS Larger volumes of distribution were observed for both aztreonam and imipenem in trauma patients than in volunteers, suggesting that standard initial dosages of the antibiotics may result in lower concentrations in these critically ill patients. Both antibiotics penetrated into the sputum of most patients; however, the degree of penetration was highly variable in relation to serum concentrations.