The relationship between debrisoquine metabolic phenotype and the pharmacokinetics and pharmacodynamics of propafenone was studied in 28 patients with chronic ventricular arrhythmias (22 extensive metabolizers [EMs] and six poor metabolizers [PMs] of debrisoquine). EMs were characterized by a shorter propafenone elimination half-life (5.5 +/- 2.1 vs 17.2 +/- 8.0, p less than .001), lower average plasma concentration (Cp) (1.1 +/- 0.6 vs 2.5 +/- 0.5 ng/ml/mg daily dosage, p less than .001), and higher oral clearance (1115 +/- 1238 vs 264 +/- 48 ml/min, p less than .001). The active metabolite 5-hydroxypropafenone, assayed in 12 patients, was identified in nine of 10 EMs but in neither of the PMs. A lower incidence of central nervous system side effects was noted in EMs (14% vs 67%, p less than .01). The magnitude of QRS widening at any given propafenone Cp was greater in EMs than PMs. There was no significant difference between EMs and PMs in effective propafenone dose or frequency of antiarrhythmic response. Inhibition of debrisoquine 4-hydroxylation by propafenone was demonstrated both in vivo and in a human liver microsomal system in vitro. We conclude that propafenone is metabolized via the same cytochrome P-450 responsible for debrisoquine's 4-hydroxylation, and that its pharmacokinetics and concentration-response relationships and the incidence of central nervous system side effects are different in patients of different debrisoquine metabolic phenotype.