We have previously demonstrated that individualized dosing of etoposide (VP16) by 72-hour infusion is feasible and that the extent of leukopenia is a function of plasma concentration, pretreatment WBC (WBCp), albumin (ALB), performance status (PS), and bone marrow function (based on transfusion requirements). In the current study, 45 patients were randomized between a fixed dose of VP16 (125 mg/m2/d) versus individualized dosing to a target WBC nadir (WBCN) of 1,700/microL. The total dose was increased by an average of 22% in the latter patients (459 +/- 130 mg/m2 v 375 mg/m2, P = .002). This was associated with a decrease in both the mean WBCN (1,510 +/- 950 v 2,500 +/- 1,420/microL, P = .013) and in the variability of the WBCN (P = .039). The VP16 clearance (mL/min) was not correlated with body surface area. Partial responses were observed in one patient each with hepatoma and non-Hodgkin's lymphoma. We conclude that pharmacologically based dosing may be a means of increasing dose intensity without increasing the incidence of life-threatening toxicity due to a decrease in variability around a target WBC.