This study was undertaken to investigate the pharmacokinetics of etoposide for optimizing its oral dosage in elderly patients with non-Hodgkin's lymphoma (NHL) using the fraction of dose absorbed calculated from the data generated from first oral and intravenous doses in the same patient. Twenty-three NHL patients (ages 61-95 years) entered this study. Each received 50 mg/m2 of etoposide by 1-hour i.v. infusion, which was repeated every 24 hours for 5 days. The second cycle commenced on day 21, with etoposide being administered by mouth at a dose as close to 50 mg/m2 as possible. Serial blood samples were collected and analyzed for etoposide by HPLC. The fraction of dose absorbed (F) was calculated as F = (AUCor/AUCi.v.) (Di.v./Dor), and etoposide was then given orally for the following 20 days at a daily dose equivalent to Dor/F. After 1 week free of etoposide administration, a second cycle of oral etoposide at the adjusted dose was given for 21 days. The mean +/- SD values for t1/2 beta, tmax, Cmax, CLTor, and MRT observed following the first oral dose were 8.98 +/- 4.84 h, 1.39 +/- 0.96 h, 0.083 +/- 0.046 mg.L-1/mg.m-2, 1.89 +/- 1.2 L.h-1/m2, and 10.37 +/- 2.76 h, respectively, and those observed following the first intravenous dose were 8.05 +/- 5.11 h, 1.57 +/- 0.17 h, 0.142 +/- 0.043 mg.L-1/mg.m-2, 1.25 +/- 0.44 L.h-1/m2, and 7.69 +/- 1.53 h, respectively. The mean +/- SD of F was 0.80 +/- 0.34. The data obtained indicate that optimization of etoposide oral dosage using F yielded good clinical results while keeping the morbidity at a level that is similar to that of the i.v. administration.