The population pharmacokinetics of cyclosporine (CsA) in adult recipients of cardiac transplants were determined from sparse, retrospective drug monitoring data accumulated for at least 3 months after surgery. All were receiving oral CsA twice daily, and morning trough levels in whole-blood were measured by high-performance liquid chromatography. Additional data included height, weight, gender, age, ethnicity, hematocrit, total bilirubin, and concurrent drug use. Population modeling was performed using NONMEM on 36 randomly selected patients, assuming a one-compartment model with first-order absorption and elimination. Improved fits were obtained by incorporating the following expression in the model to adjust oral bioavailability as a function of postoperative day (POD): F = 0.2 + 10 x ABS (POD - 7)/([POD + 10] x 60). Interpatient variability (CV%) in clearance (CL) was 20.2%. There was a mean bias of 8.5% at the average CsA concentration of 250 ng/ml when the predictive performance was assessed statistically in a reserved subset of 33 patients who received cardiac transplants. For the entire population (n = 69 patients), the average CsA CL and terminal half-life (T1/2) were, respectively: CL (l/h) = 0.256 x weight (kg); T1/2 = 11.0 hours, or CL (l/h) = 0.184 x weight (kg); T1/2 = 14.7 hours, if there was concomitant diltiazem administration. These results compared favorably with those reported elsewhere for studies of postcardiac transplant kinetics using the traditional multiple blood sampling approach.