A total of 9 patients entered in a phase I trial who received oral idarubicin daily for 3 days took part in pharmacokinetic studies, and bioavailability studies were performed on 13 additional patients receiving single doses of oral idarubicin alternating with i.v. treatment. The data were best fit by a two-compartment model (distribution and elimination compartments for i.v. drug and absorption and single-phase elimination for oral drug). For different idarubicin doses in the phase I and bioavailability studies, the median values for the terminal half-life of idarubicin varied from 5.6 to 11.6 h. High concentrations of the active metabolite idarubicinol were formed. Idarubicinol was eliminated more slowly than was the parent compound, with median half-lives for different dose levels varying from 8 to 32.7 h. Although most pharmacokinetic parameters were similar in plasma and whole blood, peak concentrations and AUCs in whole blood were about 3-4 times those calculated in plasma for idarubicin and about 1.5-2 times those determined in plasma for idarubicinol, indicating fairly extensive uptake into erythrocytes. Oral bioavailability was determined by comparing oral idarubicin to i.v. drug with respect to the combined idarubicin and idarubicinol plasma AUCs, and it varied from 12%-49% (median, 29%). Bioavailability was essentially the same (30%) when whole-blood values were used. Urinary excretion of the drug was less than 5% of the delivered dose by 96 h. Granulocytopenia correlated with plasma idarubicinol "estimated" clearance and steady-state volume of distribution, with whole-blood idarubicinol AUC, area under the moment curve (AuMC), and "estimated" clearance and volume of distribution, and with whole-blood combined idarubicin and idarubicinol AUCs. This suggests that drug contained in erythrocytes plays a major role in toxicity and that idarubicinol may play a larger role in toxicity than does the parent compound.