1 In the present study, responses of human omental small arteries and veins to endothelin-1 and endothelin-3 were characterized by use of the ETB receptor selective agonist, sarafotoxin S6c, the ETA receptor antagonist, BQ123, the ETB receptor antagonist, IRL1038, the NO-synthase inhibitor NGmonomethyl-L-arginine (L-NMMA, 300ELM) and indomethacin (10I1M).2 Small arteries (internal diameter 413 ± 22 Am) and parallel running veins (646 ± 35 Am) were mounted in a myograph under a normalized tension equivalent to 90% of a transmural pressure of 100 mmHg and 19 mmHg in vivo, respectively. 3 In small arteries and veins, endothelin-1 caused a concentration-dependent increase in wall tension (E.. = 3.90 ± 0.56 mN mm-' and 1.90 m ± 0.32 mN mm-' respectively, P < 0.05) and was equipotent (arteries: pD2 = 8.91 ± 0.11; veins: pD2 = 8.63 ± 0.08, NS). In endothelium intact arteries, L-NMMA significantly enhanced the sensitivity to endothelin-1 (pD2 control: 8.92 ± 0.16; pD2 L-NMMA: 9.37 ± 0.11; P<0.05). L-NMMA did not affect the sensitivity of veins to endothelin-1. Indomethacin was without effect in arteries and veins. In veins, endothelin-3 was about a hundred times less potent than endothelin-l and showed a biphasic response curve. Small arteries did not contract to endothelin-3. Neither small arteries nor veins contracted to sarafotoxin S6c. Furthermore, no relaxation to endothelin-1 or sarafotoxin S6c was seen in any precontracted vessels. 4 BQ123 (0.03-3 3LM) produced a concentration-dependent rightward parallel displacement of the endothelin-l concentration-response curve in small arteries and veins yielding pA2 values of 7.09 and 7.48 respectively. The slope of the Schild plot in arteries and veins was 1.26 ± 0.24 (NS from unity) and 0.61 ± 0.13 (P <0.05 compared to unity) respectively. IRL1038 (3 JLM) did not affect the potency of endothelin-1 in arteries and veins. In veins, the low sensitivity component (pD2 = 7.16 ± 0.08) of the biphasic response curve to endothelin-3 was completely blocked by BQ123 (31JM), whereas the high sensitivity component (pD2 = 8.66 ± 0.08) was resistant to BQ123 (3 JiM) and IRL1038 (3 JAM).5 These results indicate that contractions of human small vessels to endothelin-l are predominantly mediated by ETA receptors and that nitric oxide modulates the response to endothelin-l in small arteries but not in veins. The different antagonistic potency of BQ123 against endothelin-l and the differential endothelin-1/endothelin-3 potency ratios in arteries and veins provide evidence for the hypothesis that ETA receptors in human small arteries are different from ETA receptors in human small veins. There is no evidence of contractions mediated by 'classical' ETB receptors in these vessels, but small veins appear to contain a functional non ETA/non ETB receptor with a high affinity for endothelin-3.