Methotrexate (MTX) and 7-hydroxy-methotrexate (7-OH-MTX) plasma concentration-time curves (AUC) have been analyzed in 24 patients after different routes of MTX administration. After an i.v. bolus (50 mg/m2), the AUC for 7-OH-MTX is correlated with that for MTX and inversely correlated with the MTX plasma clearance. When MTX is administered with plasma steady level standardization, using the test-dose protocol, at a level of 10(-5) M over 36 hr (10(-5), 36 hr), 7-OH-MTX-AUC is still correlated with the i.v. bolus pharmacokinetic parameters. The dose prediction using the classical test-dose protocol provides a less efficient MTX dose adjustment at 5 X 10(-4) M over 8 hr (5.10(-4), 8 hr) and the hydroxylation process is no more correlated with the i.v. parameters. On the opposite, upon 6 successive infusions with 10(-5), 36 hr or 5.10(-4), 8 hr protocols, the plasma concentrations of 7-OH-MTX are not significantly modified. This suggests that the hydroxylation process is not inducible.