The area under the plasma concentration-time curve (AUC) following an intravenous dose of etoposide varies considerably among patients, which in part contributes to the unpredictability of toxicity and response seen in individual patients. This study evaluated the utility of therapeutic drug monitoring of etoposide in reducing the interpatient variability of etoposide steady-state concentrations during prolonged infusion. The etoposide prodrug etoposide phosphate (Etopophos; Bristol-Myers Squibb Company, Princeton, NJ) was administered by infusion using an adaptive dosing strategy. It was given in combination with carboplatin, which was dosed on an AUC basis. Patients with histologically or cytologically proven small cell lung cancer were treated with etoposide phosphate by continuous 120-hour infusion and carboplatin at a dose calculated by the Calvert formula to give an AUC of 5 mg/ mL.min. Blood samples were taken on days 2 through 5 of each treatment cycle, and high-performance liquid chromatography was used to measure the plasma etoposide concentration. The resultant concentrations were compared with target concentrations of 1 or 2 micrograms/mL and these data were used to calculate the rate of infusion for the following 24 hours. In the first cohort, the target etoposide concentration was 1 microgram/mL, and this was achieved (mean +/- SD = 1.05 +/- 0.24 micrograms/mL) by infusing etoposide phosphate doses of 21 to 109 mg (15 to 68 mg/m2) per day. In the second cohort, the target concentration of 2.0 micrograms/mL was achieved (mean +/- SD = 2.05 +/- 0.31 micrograms/mL) with infused etoposide phosphate doses of 69 to 193 mg (41 to 114 mg/m2) per day. This technique reduced the variation in plasma levels and resulted in predictable hematologic toxicity. Cumulative hematologic toxicity necessitated an extension of the treatment cycle from 3 to 4 weeks, however. Of six evaluable patients, two had a complete response and one had a partial response. Therapeutic drug monitoring was shown to reduce the interpatient variation in the plasma etoposide concentration by half and shows promise for individualizing treatment with combination chemotherapy. Exploiting the known relationships between the pharmacokinetics and pharmacodynamics of these two drugs by using therapeutic drug monitoring may lead to better therapeutic safety and efficacy.