1 The disposition in urine of debrisoquine and its hydroxylated metabolites has been studied in subjects of the 'extensive metabolizer' (EM; n = 5) and 'poor metabolizer' (PM; n = 5) phenotypes. The 4-hydroxylation of debrisoquine by PM subjects following a 10 mg oral dose was capacity-limited and displayed significant dose-dependency over a range of 1-20 mg. In contrast, the EM subjects' ability to perform this metabolic oxidation did not deviate from first-order kinetics over a dose range of 10-40 mg. 2 The disposition of debrisoquine in plasma following a 10 mg oral dose has been studied in EM (n = 4) and PM (n = 3) subjects. Whilst PM subjects displayed significantly higher plasma levels of debrisoquine at all time points following 1 h post-dosing, and higher values for areas under the plasma concentration-time curve (EM: 105.6 +/- 7.0 ng ml-1 h; PM: 371.4 +/- 22.4 ng ml-1 h, 2P less than 0.0001), neither debrisoquine plasma half-life (EM: 3.0 +/- 0.5 h; PM: 3.3 +/- 0.4 h) nor renal clearance of the drug (EM: 152.8 +/- 30.3 ml min-1; PM: 137 +/- 4.5 ml min-1) displayed significant inter-phenotype differences. 3 The results of these investigations show that the phenotyping of individuals for debrisoquine oxidation status by means of a 'metabolic ratio' derived from a single 0-8 h urine sample has a sound kinetic basis. The kinetic differences between the two phenotypes would strongly suggest that the metabolic defect manifested in PM subjects is one of pre-systemic elimination capacity.