Eight healthy subjects were administered trazodone-HCl orally (100 mg) with and without food and by infusion in a three way cross-over study. Unchanged trazodone was determined in serum and urine by high performance liquid chromatography after an alkaline extraction. Absorption of trazodone was irregular in fasting subjects and improved after food intake. Food intake significantly decreased the maximum serum concentrations of trazodone from 1.88 +/- 0.42 to 1.47 +/- 0.16 micrograms/ml, and increased the time for reaching maximum concentration from 1.3 +/- 0.8 hr to 2.0 +/- 1.5 hr. No differences were observed in the total amount of trazodone absorbed with or without food with bioavailability values of 65 +/- 6 and 63 +/- 4 per cent, respectively. The apparent volume of distribution and total body clearance for trazodone were estimated to 0.84 +/- 0.16 l/kg and 5.3 +/- 0.9 l/hr, respectively. The terminal elimination half-life of 7.3 +/- 0.8 hr showed no significant differences between the different ways of administration. Urinary excretion of unchanged trazodone during 26 hr was less than 0.13 per cent of the administered dose, suggesting a high degree of trazodone metabolism. Earlier statements of enterohepatic circulation of trazodone and pharmacokinetic differences between males and females were not confirmed by the present study. Due to the irregular absorption in fasting subjects, trazodone should preferably be administered after food.