1. In vitro experiments in a microvascular myograph were designed to characterize postjunctional muscarinic receptors producing contraction both in the presence and absence of the endothelium in coronary resistance arteries (normalized diameter of 150-450 microns), isolated from the left ventricle of hearts from 3-6 month old lambs. Preferential muscarinic receptor antagonists were used to determine the receptor subtype: pirenzepine (M1 receptor), AFDX 116 (M2 receptor), 4-DAMP and pFHHSiD (M3 receptor). 2. The rank order of potency for muscarinic agonist-induced increases in tension in endothelium-intact preparations was oxotremorine-M = methacholine = acetylcholine (ACh) > carbachol. Removal of the endothelium increased the potency of ACh, but this procedure did not change either the sensitivity or maximal response to carbachol. 3. The contractile response to ACh was reproducible. Incubation with 3 x 10(-7)-3 x 10(-6) M pirenzepine induced non-parallel rightward shifts and depressed the maximum of the concentration-response curve to ACh in endothelium-intact arteries. The slope by Schild analysis was 2.9 +/- 0.8 (P < 0.05, n = 7). Atropine, AFDX 116, 4-DAMP and pFHHSiD produced parallel rightward shifts of the curves to ACh and the slopes of the Schild plots were not significantly different from unity. The pKB values for the antagonists from plots constrained to unity in endothelium-intact segments were: atropine (9.4), 4-DAMP (9.0), pFHHSiD (7.9) and AFDX 116 (6.2). 4. In endothelium-denuded arteries, pirenzepine, AFDX 116 and pFHHSiD caused concentration-dependent, parallel rightward displacements of the concentration-response curves to ACh and the slopes of the Schild plots were not significantly different from unity. The plots constrained to a slope of unity gave the following pKB values: pFHHSiD (8.7), pirenzepine (7.5) and AFDX 116 (6.2). 5. In the presence of the endothelium, low concentrations of pirenzepine (10(-9)-10(-7) M) produced leftward shifts of the ACh concentration-response curves. This potentiating effect of pirenzepine was reversed by endothelial cell removal. In preparations precontracted with the thromboxane-mimetic, U46619, the putative M1-selective agonist, McN-A-343, induced a biphasic relaxation with log IC50 of 8.53 +/- 0.14 and 5.02 +/- 0.08 for the first and second phase of the relaxation, respectively, and maximal relaxations of 22.8 +/- 4.3% and 41.1 +/- 5.4% (n = 16). McN-A-343 relaxed the vessels in the presence of 10(-7) M pFHHSiD and 3 x 10(-7) M AFDX 116, but not after incubation with 10(-9) M pirenzepine. 6. It is concluded from the pKB values for the antagonists used, that contraction induced by ACh in lamb coronary resistance arteries, in either the presence or the absence of the endothelium, is mediated via the M3 subtype of muscarinic receptors, while muscarinic receptors of another subtype at the endothelium seem to modulate the contractile response to ACh.