1. In the presence of either barium (0.2 mM) or caesium (20 mM), carbachol (3-300 microM) depolarized isolated guinea-pig ventricular myocytes. Carbachol induced an inward current under voltage clamp at a holding potential equal to the resting potential (-75 mV). 2. Acetylcholine and oxotremorine also evoked an inward current but were less effective than carbachol. Atropine (0.3 microM) prevented the depolarization and inward current induced by carbachol and acetylcholine but not by oxotremorine. Moreover, oxotremorine, but not carbachol, induced an inward current in the absence of extracellular sodium. 3. Carbachol increased membrane chord conductance when it induced an inward current. These effects were recorded under experimental conditions that suppressed the voltage- and time-dependent sodium current (tetrodotoxin) and calcium current (cadmium), the inwardly rectifying potassium current, iK1 (caesium, barium and tetraethylammonium) and the current generated by the sodium-potassium pump (zero external potassium). 4. Under these same experimental conditions, the steady-state I-V relationship in the presence of carbachol was subtracted from that in its absence. The apparent reversal potential (Erev) was 25 mV with extracellular Na+ ([ Na+]o) at 143 mM and intracellular Na+ ([Na+]i) at 11 mM. Replacement of [Na+]o by N-methyl-D-glucamine was associated with a shift of the apparent Erev to more negative voltages by approximately 61 mV per tenfold change of [Na+]o. 5. Isoprenaline induced an inward current in ventricular myocytes that depended upon sodium entry, required the accumulation of cyclic AMP and which was partially suppressed by acetylcholine (Egan, Noble, Noble, Powell, Twist & Yamaoka, 1988). In contrast to the current evoked by beta-adrenoceptor agonist, the current induced by muscarinic agonist was smaller and sustained. Moreover, the carbachol-induced current was not suppressed by prior addition of isoprenaline. 6. The findings are consistent with the mechanism that carbachol activates a plasma membrane ion channel that admits sodium and thereby increases intracellular sodium activity. The estimated increase of intracellular sodium activity from electrophysiological data agrees quantitatively with that obtained from measurements with sodium-sensitive microelectrodes (Korth & Kühlkamp, 1985). 7. The ability of carbachol to increase sodium influx may be the first step in a series of reactions that eventually alters sodium-calcium exchange and could account for catecholamine-independent stimulation of developed force in mammalian ventricle.