A three-phase trial has been done in 11 volunteers. They were given 600 mg phenytoin (Dilantin capsules) in each phase after an overnight fast. In the first study, phenytoin was given alone. In the second phase 400 mg carbamazepine (CBZ) was given at the same time as the phenytoin, and in the third part, 200 mg CBZ t.d.s. was given for one week prior to the phenytoin. Blood samples were taken for 72 h in each phase. Plasma levels of phenytoin and CBZ were determined by HPLC, and plasma protein binding was determined by equilibrium dialysis. The unbound fraction of phenytoin was 0.082, 0.085, and 0.077 in the control, single-dose CBZ, multi-dose CBZ phases, respectively. Single and multiple doses of CBZ decreased the plasma level of phenytoin. The 72-h AUC of phenytoin was 276, 237, and 176 mg h.l-1 in the 3 phases, respectively, and the 72-h AUC of unbound phenytoin was 22.8, 20.5, 13.0 mg h.l-1. The AUC of phenytoin (unbound and total) after multiple doses of CBZ was significantly lower than in the other two phases. The apparent volume of distribution (Vz/f) was 89.9, 110.3, and 121.3 l in the 3 phases, respectively. Through pharmacokinetic analyses, the decreased AUC and increased Vz/f were attributed to decreased bioavailability of phenytoin when CBZ was co-administered.