Chloramphenicol sodium succinate (SCAP) kinetics were studied in 10 critically ill patients. High-performance liquid chromatography was used to assay SCAP and chloramphenicol (CAP, antibiotic) in serum and urine. Total body (ClTB), metabolic (ClM) and renal (ClR) clearances of SCAP were variable. Correlations were found between creatinine clearance (Clcr) and ClTB, ClM and ClR of SCAP (r [correlation coefficient] = 0.92, P < 0.001; r = 0.84, P < 0.005; and r = 0.84, P < 0.005). Recovery of SCAP in the urine demonstrated large interpatient variability. Between 6.5% and 43.5% of the SCAP dose was recovered in the urine of 6 patients. This variability could not be explained by incomplete urine collection or by differences in renal function. Renal excretion of SCAP influenced CAP serum levels. CAP ClTB was diminished, but no relationship was found between routine liver function studies and CAP ClTB. Such relationships should not be used to monitor CAP in critically ill patients.