The objective of this study was to develop a limited sampling model (LSM) to estimate the area under the curve (AUC) of 7-ethyl-10-[4-(1-piperidino)-1-piperidino]carbonyloxycamptothecin (CPT-11) and that of 7-ethyl-10-hydroxycamptothecin (SN-38) as predictive pharmacokinetic variables for leukopenia and episodes of diarrhea induced by CPT-11 administration. The model was developed with a training set consisting of pharmacokinetic studies in 36 patients who received a 90-min i.v. infusion of CPT-11 at a dose of 100 mg/m2. A multiple regression analysis of CPT-11 or SN-38 concentrations observed at each time point in the training set was used to predict the AUC of CPT-11 or SN-38. The final sampling models using only two time points were: AUCCPT-11 = 3.7891*C2.5 + 14.0479*C13.5 + 1.5463 AUCSN-38 = 0.5319*C2.5 + 19.1468*C13.5 + 72.7349 where C2.5 and C13.5 are the plasma concentration of CPT-11 (micrograms/ml) or SN-38 (ng/ml) at 2.5 and 13.5 h after the initiation of CPT-11 infusion, respectively. The models were validated prospectively on a separate test data set of 12 patients receiving the same dose of CPT-11 investigated in a previous study. Validation of the final LSM on the test data set gave values of root mean square error (RMSE) of 12.72% and 5.97% for the AUC of CPT-11 and that of SN-38, respectively. The model can be used to monitor the AUCs of both CPT-11 and SN-38 for the early prediction of toxicities and to establish a pharmacokinetically based dose modification strategy for safe administration of CPT-11.