Five different subtypes (human m1, m2, m5 and rat m3, m4) of muscarinic acetylcholine receptors (mAChR) were produced in insect Sf9 cells by infection with recombinant baculoviruses. N-[3H]methylscopolamine ([3H]NMS) has a similar affinity to each of these mAChR subtypes in cell membranes, while pirenzepine, 11-((2-[(diethylamino)methyl]-1-piperidinyl)acetyl)-5,11-dihydro-6H-pyri do- (2,3-b)(1,4)benzo-diazepin-6-on (AF-DX 116) and (+/-)-p-fluoro-hexahydrosiladifenidol (p-F-HHSiD) have a higher affinity for m1, m2 and m3, respectively, than for the other subtypes, indicating the maintenance of subtype specificity of mAChR in this system. Digitonin (1%, w/w) with sodium cholate (0.1%, w/w) solubilized 51% of m1, 36% of m2, 3% of m3, 28% of m4 and 17% of m5 mAChR from these cell membranes with retention of the [3H]NMS binding activity. Optimization of cholate concentrations resulted in solubilization of up to 50-60% for m1, m2 and m4, but up to 25% for m5 and 7% for m3. Optimal concentrations of cholate differed from one subtype to another. Sucrose monolaurate solubilized 21-43% of m1, m2 and m4, but only up to 12% for m5 and 2% for m3. 3-(3-cholamidopropyl)dimethylammonio-1-propanesulfonate (CHAPS) was practically ineffective in mAChR solubilization from Sf9 cell membranes for all subtypes investigated. Solubilization with digitonin and cholate had little influence on [3H]NMS affinity for m2 and m4, but decreased m1 and m5 affinity by 10-fold and that of m3 by more than 50-fold. These results indicate that the solubility and stability of mAChR in detergents differ among the subtypes, in spite of their structural similarities. These differences should be taken into account when comparing the five subtypes, particularly when determining the proportion of each subtype in a given tissue by precipitating the solubilized mAChR with subtype-specific antibodies.