The plasma concentration-time profiles of meperidine following intravenous injection in surgical patients and volunteers were investigated by reference to a classical two-compartment open model. Physiologic characteristics of the subject and variables associated with the surgery and anesthesia were screened as determinants of the kinetic patterns observed. When meperidine administration preceded induction of anesthesia, induction was consistently followed by an increase in venous plasma concentrations that prevented classical kinetic analysis. To facilitate calculations in subsequent studies in patients, meperidine injections were made following induction of anesthesia. Type of anesthesia or premedication, patients' sex, or cigarette smoking did not appear to be important factors in this evaluation. Increasing alcohol consumption was associated with increasing volumes of distribution. Increasing age was associated with increasing fraction of drug unbound in plasma. These factors may relate directly to clinical observations that heavy alcohol consumers tend to be more refractory to central nervous system (CNS) depressants and that elderly patients are more susceptible to respiratory depression from narcotics.