1. Accurately predicting the kinetics in man greatly improves the design of the phase I clinical studies. This was particularly crucial in the case of bosentan, a new endothelin receptor antagonist, as very large interspecies differences in systemic clearance were observed in the animal species investigated, namely from 1.5ml/min/kg in the dog up to 70 ml/min/kg for the rabbit. 2. Bosentan was shown to be metabolized by the hepatic cytochrome P450, therefore the rate of metabolism was investigated in vitro in liver microsomes and hepatocytes, across the species which had been tested in vivo. The same rank-order of metabolism was found for the laboratory animals both in vitro and in vivo, and hepatocytes appeared to be more representative of the in vivo situation than liver microsomes. The in vitro clearance in human hepatocytes was very close to that observed in dog hepatocytes. 3. A plasma clearance for bosentan in man of 1-2 ml/min/kg was predicted by combining the in vivo and in vitro data from a few animal species with the in vitro data in man. This expectation was subsequently found to agree reasonably well with the plasma clearance observed in healthy volunteers: ca 2 ml/min/kg. Integrating this prediction into the design of the first clinical protocols substantially improved the quality of the human pharmacokinetic data obtained.