1 The effects of acetylcholine and substance P on the efflux of Rb-86+ and K-42+ from rat aorta and pig coronary artery, respectively, were compared with those of the K+ channel opening agent, cromakalim. 2 In rat aorta preloaded with Rb-86+ and/or K-42+, acetylcholine produced transient, concentration-dependent increases in the efflux rate coefficients of these tracers (maximum almost-equal-to 35%). These effects were abolished by endothelial cell removal. 3 Donor/acceptor experiments with rat aorta suggested that at least some of the efflux of Rb-86+ seen in the presence of acetylcholine was not derived from the endothelium, but came from the smooth muscle itself. 4 Acetylcholine (10-mu-M)-induced Rb-86+ efflux was reduced by tetraethylammonium (TEA, 10 mM) to 33% and ouabain (300-mu-M) to 54% of control. Preincubation with Ba2+ (100-mu-M) did not significantly inhibit acetylcholine-induced efflux. 5 Acetylcholine-induced K-42+/Rb-86+ efflux was unaffected by preincubation with glibenclamide (10-mu-M). In contrast, the K-42+/Rb-86+ efflux induced by cromakalim was inhibited by glibenclamide (50 nM) by 50%. 6 Acetylcholine (0.3-10-mu-M)-induced inhibition of phenylphrine (1-mu-M)-induced tone was abolished by endothelial cell removal but unaffected by glibenclamide. Cromakalim-induced relaxations were endothelium-independent and were inhibited by glibenclamide in a concentration-dependent manner. 7 L(G)-monomethyl L-arginine (L-NMMA, 250-mu-M) produced a significant (37 +/- 14%) inhibition of acetylcholine-induced Rb-86+ efflux whereas D(G)-monomethyl L-arginine was without effect. In the tissue bath L-NMMA inhibited relaxations produced by acetylcholine (0.3-10-mu-M), but was without effect on responses to cromakalim. 8 In the pig coronary artery, substance P induced an endothelium-dependent efflux of Rb-86+ and K-42+, which was unaffected by preincubation with glibenclamide (10-mu-M) or L-NMMA (250-mu-M). 9 The present study shows that acetylcholine and substance P each open K+-channels in arterial smooth muscle. However, the insensitivity of the stimulated Rb-86+ /K-42+ efflux to inhibition by glibenclamide suggests that the K+-channel opened by these agents is different from the K+-channel opened by cromakalim. In addition, the inability of L-NMMA to inhibit fully the acetylcholine- and substance P-stimulated Rb-86+ efflux suggests that in rat aorta and pig coronary artery Rb-86+ efflux suggests that in rat aorta and pig coronary artery the endothelium-derived hyperpolarizing factor(s) (EDHF) is different from endothelium-derived relaxing factor (EDRF).