The pharmacokinetics of dexfenfluramine (d-F) and its metabolite dexnorfenfluramine (d-NF) were compared in 10 obese (145 +/- 13 s.d. % of ideal body weight (IBW)) and 10 non-obese healthy volunteers (93 +/- 8% IBW). Each group included five men and five women, aged 28 +/- 8 years. Subjects were given single doses of d-F i.v. (15.5 mg base infused over 3 h) and orally (25.9 mg base in capsules) on separate occasions. After i.v. infusion in obese subjects, the volume of distribution (Vss) of d-F was significantly higher (969.7 +/- 393.3 l; 95% CI 688.6-1250 l) than in controls (668.7 +/- 139.6 l; 95% CI 568.9-768.5 l; P < 0.01). Clearance was not significantly different (43.9 +/- 21.0 l h-1 vs 37.3 +/- 10.6 l h-1) and the terminal half-life tended to be longer (17.8 +/- 9.4 vs 13.5 +/- 3.9 h NS). Combined data from the two groups indicated a positive correlation between Vss and % IBW (r = 0.544; P < 0.02). The oral bioavailability of d-F was 0.61 +/- 0.15 in obese subjects and 0.69 +/- 0.11 in controls. There was no significant difference between obese subjects and controls in Cmax, tmax and t1/2,z (Cmax: 20.1 +/- 6.7 and 27.3 +/- 6.2 micrograms l-1; tmax: 3.5 vs 3.0; t1/2,z: 16.5 +/- 7.1 vs 14.5 +/- 2.6 h respectively). The AUC ratio expressed in molar units for d-F/d-NF was 2.29 +/- 1.78 (i.v.) vs 1.25 +/- 0.64 (oral) in obese subjects and 2.05 +/- 1.26 (i.v.) vs 1.40 +/- 0.87 (oral) in controls.(ABSTRACT TRUNCATED AT 250 WORDS)