Membrane potential and current responses to the muscarinic agonist carbachol (CCh) were investigated in single ileal smooth muscle cells isolated from humanely killed M₂ or M₃ knockout (KO) mice as well as wild-type (WT) mice. Membrane potential responses were recorded using the nystatin-perforated current-clamp technique, and membrane current responses using the conventional whole-cell voltage-clamp technique. Data are expressed as mean ± S.E.M. Unpaired Student’s t test was used for statistical comparison. In cells from M₂ KO mouse, 1 μM CCh produced a sustained depolarization, on which an increased frequency of action potential discharge was often superimposed. The size of depolarizations evoked had a mean amplitude of 12.2 ± 1.2 mV (n=4), much smaller than that in WT cells (41.8 ± 1.5 mV, n=4, P<0.05). In M₃ KO cells, CCh at 1 μM was almost without effect (n=5) and even at 100 μM produced no more than 10 mV depolarization (n=4). In M₂ KO cells bathed in a Na⁺-rich physiological medium and dialysed intracellularly with a CsCl-based solution, application of CCh (100 μM) under voltage-clamp at -50 mV evoked a biphasic inward current consisting of an initial transient followed by a sustained component, as seen in WT. However, the respective components (67.5 ± 28.3 pA and 3.7 ± 1.0 pA, n=10) were significantly smaller than those in the WT (161.9 ± 26.6 pA and 23.1 ± 6.2 pA, n=11, respectively, P<0.05). Replacement of the intracellular Cl^– with glutamate^– (n=6) or cell treatment with the Cl^– channel blocker niflumic acid (n=3), resulted in abolition of the transient component alone. In M₃ KO cells, CCh (100 μM) evoked a slight sustained inward current (4.0 ± 0.4 pA, n=7), and the current response become more clear (11.5 ± 1.5 pA, n=15) if the extracellular Na⁺ was replaced with Cs⁺. Non-stationary noise analysis of CCh (100 μM)-evoked currents revealed that the unitary channel conductances underlying the sustained inward currents in M₂ KO and M₃ KO cells were 10.8 ± 1.5 pS (n=8) and 2.2 ± 0.2 pS (n=8), respectively. These values clearly differed from the corresponding value (40.1 ± 4.4 pS, n=6, P<0.05) estimated for the WT. These results suggest that M₂ receptor stimulation depolarizes the cell via activating 2-pS cationic channels, and M₃ receptor stimulation does so via activating both 10-pS cationic channels and Ca²⁺-activated Cl^– channels. In intact cells, beside these channels, the 40-pS cationic channels which are activated synergistically by both M₂ and M₃ subtypes may play a crucial role in the muscarinic depolarization.