1. ETA and ETB-selective and non-selective ligands were used to define the endothelin receptors in the media (vascular smooth muscle layer) of human aorta and coronary artery. Saturation experiments with iodinated endothelin-1 (ET-1), endothelin-2 and sarafotoxin 6b (S6b) identified high affinity binding sites in aorta (KD [125I]-ET-1 0.33 +/- 0.02 nM (n = 9), KD [125I]-ET-2 1.04 +/- 0.23 nM (n = 5), KD [125I]-S6b 0.15 +/- 0.01 nM (n = 9 +/- s.e.mean)) and coronary artery (KD [125I]-ET-1 0.43 +/- 0.10 nM, KD [125I]-ET-2 0.71 +/- 0.17 nM, KD [125I]-S6b 0.27 +/- 0.03 nM (n = 3 +/- s.e.mean)). Hill coefficients (nH) approached unity in each case. 2. No specific binding was detectable with [125I]-ET-3 (4 pM-4 nM) in aorta. Unlabelled ET-3 competed monophasically with [125I]-ET-1 in aorta (KD, 8.21 +/- 1.62 nM, compared to unlabelled ET-1 KD, 0.60 +/- 0.20 nM) (n = 3 +/- s.e.mean). In coronary artery, the KD and Bmax values calculated from [125I]-ET-3 saturation experiments were 2.13 +/- 1.39 nM and 20.6 +/- 12.9 fmol mg-1 protein, respectively (n = 3 +/- s.e.mean). 3. ETA antagonists competed monophasically for [125I]-ET-1 (100 pM) binding sites with nanomolar or subnanomolar affinity in the aorta (KD BQ123, 0.47 +/- 0.13 nM; KD FR139317, 0.40 +/- 0.10 nM; KD PD151242, 2.09 +/- 0.48 nM) and coronary artery (KD FR139317, 0.41 +/- 0.13 nM; KD PD151242, 3.60 +/- 0.74 nM) (n = 3 +/- s.e.mean). However, two site fits were preferred on analysis of competition experiments with ETB-selective agonists versus [125I]-ET-1 in coronary artery (BQ3020: KDETA 0.96 +/- 0.14 microM, KD ETB 1.34 +/- 1.08 nM and sarafotoxin 6c: KD ETA 1.15 +/- 0.14 microM, KD ETB 1.77 +/- 0.72 nM) (n = 3 +/- s.e.mean). The selectivity of the agonists for ETB receptors (700 fold) was lower than reported in other species. 4. Sarafotoxin 6b (2 pM-2 microM) completely inhibited [125I]-ET-1 (100 pM) binding in aorta (KD 1.36 +/- 0.22 nM) (n = 3 +/- s.e.mean). The non-peptide compounds Ro462005 and bosentan, competed with [125I]-ET-1 binding in coronary artery with KD values of 0.19 +/- 0.04 microM and 2.94 +/- 0.95 nM, respectively (n = 3 +/- s.e.mean). 5. Inhibition of [125I]-ET-2 and [125I]-S6b binding by FR139317 was similar to the inhibition of [125I]-ET-1 binding in both arteries, being monophasic with KD values in the same range. 6. ETA receptors in coronary artery media were detected by [125I]-PD151242 (KD 0.23 +/- 0.04 nM, Bmax 10.1 +/- 1.2 fmol mg-1 protein) (n = 3 +/- s.e.mean). [125I]-BQ3020, an ETB-selective radioligand, indicated the presence of a smaller population of ETB receptors in this tissue (KD 0.60 +/- 0.31 nM, Bmax 4.5 +/- 2.1 fmol mg-1 protein) (n = 3 +/- s.e.mean). 7. Autoradiography with [125I]-PD151242 and [125I]-BQ3020 confirmed the predominance of ETA receptors in the media of both arteries. 8. The results of this study indicate that ETA receptors predominate in the vascular smooth muscle of human cardiac arteries, with a small and variable population of ETB receptors detectable in the coronary artery.