1. Eight healthy Thai males, aged 19-27 years, received quinine or quinidine dihydrochloride 10 mg kg-1 body weight by intravenous infusion over 1 h. At least 1 week later, the alternative alkaloid was administered. 2. The terminal elimination half-time of quinidine was shorter than that of quinine (median [range]; 5.7 [5.0-10.0] vs 9.9 [8.8-15.1] h, P < 0.01), the volume of distribution at steady state (Vss) for quinidine was larger than that for quinine (3.5 [2.5-5.6] vs 3.1 [1.8-4.1] 1 kg-1; P = 0.02) and quinidine was less bound to plasma proteins (% free drug: 22.8 [15.4-47.2] vs 9.4 [7.3-15.0]%, P < 0.01). Total clearance was greater for quinidine (7.7 [3.9-11.4] vs 3.4 [1.8-4.6] ml min-1 kg-1, P < 0.01) but not for clearance of unbound drug (32.2 [14.6-50.4] vs 29.9 [20.2-50.9] ml min-1 kg-1 respectively, P > 0.2). 3. Side-effects, including transient hypotension after quinidine in two cases, were mild. 4. Both drugs produced prolongation of the rate-corrected QT interval (QTc), with similar rates of elimination from the cardiac conduction 'effect' compartment (keo; 4.14 [0.03-15.33] h-1 for quinine, 3.74 [1.63-13.14] h-1 for quinidine, P > 0.19). Using a linear concentration-response model, the intercept ('threshold') for quinidine effect was lower than that for quinine (P = 0.004) but the slopes (change in QTc for a given change in free drug concentration) were similar (P = 0.56).(ABSTRACT TRUNCATED AT 250 WORDS)