We have determined the affinity and selectivity of a new nonpeptide antagonist PD156707 (sodium 2-benzo(1,3ioxol-5-yl-4-(4-methoxy-pheny l)-4-oxo-3-(3,4,5-trime tho xybenzyl)-but-2-enoate) for human endothelin (ET)(A) and ET(B) receptors. In human coronary artery and saphenous vein the affinity of the ET(A) receptor for PD156707 was 0.15 +/- 0.06 nM and 0.5 +/- 0.13 nM, respectively. Competition experiments in human left ventricle and kidney revealed that PD156707 had 1,000- to 15,000-fold selectivity for the ET(A) receptor over the ET(B) receptor. This selectivity was confirmed autoradiographically. In human coronary artery, mammary artery and saphenous vein PD156707 (3-300 nM) potently antagonized the vasoconstrictor responses to ET-1. The pA2 values estimated from the Gaddum-Schild equation were 8.07 +/- 0.09, 8.45 +/- 0.11 and 8.70 +/- 0.13, respectively. The concentration-response curves to ET-1 were shifted to the right in parallel fashion, without reduction of the maximum response. However, the regression lines fitted to the resulting Schild data deviated significantly from one. PD156707 appeared to be a more effective antagonist at lower concentrations than at the higher ones. It is possible that PD156707, a sodium salt, was reverting to a less soluble form which results in underestimation of its potency. These data show that PD156707 is a potent and selective antagonist at human ET(A) receptors and will be useful in clarifying the role of the endothelin peptides in human cardiovascular disease.