Antitumor activity of topotecan, a new derivative of camptothecin, has been documented in many tumors. The active lactone form of topotecan is in equilibrium with the inactive hydroxyacid form. However, dose-limiting toxicity, neutropenia, is correlated to the area under the concentration time curve (AUC) of not only the lactone form but also total (lactone + hydroxyacid) topotecan. Because the determination of the total topotecan plasma concentration is technically much easier than the lactone, we sought to establish a limited sampling model for the AUC of total topotecan. Thirty-four pharmacokinetic profiles were obtained in 19 patients in a Phase I study of topotecan, which was infused over 30 min for 5 days. Multiple regression models predicting the AUC were developed using 17 profiles and validated using the rest of the data. The best model was: AUCpred (ng x h/ml) = 1.75 x C15m (ng/ml) + 11.2 x C6h (ng/ml) + 7.90 x dose (mg/m2), where AUCpred was the AUC predicted by the model, and C15m and C6h were the measured concentrations of total topotecan at 15 min and 6 h after the end of infusion, respectively. When this model was validated, it was unbiased (percentage of mean predicted error +/- SE, -1.0 +/- 3.3%) and precise (percentage of root mean square error, 11%). Because this model requires only two concentrations of total topotecan to estimate the AUC, it will be useful for further pharmacodynamic evaluation of topotecan in multi-institutional studies.