Ethanol is known to cause a leftward shift of the acetylcholine concentration-response curve for channel opening of the nicotinic acetylcholine receptor (nAcChoR). However, it remains uncertain whether the mechanism underlying ethanol's effect is an increase in the binding affinity of the agonist to the receptor or an increase in the open/closed equilibrium for those receptors occupied by agonist. In the present study, this question was resolved by measuring the efflux of 86Rb+ over 9 msec from Torpedo vesicles after rapid mixing with the partial agonist suberyldicholine with or without ethanol as appropriate. Suberyldicholine's concentration-response curve is bell-shaped. Two actions underlie this bell-shaped curve, namely activation at low concentration (apparent dissociation constant for activation, Ka = 38 microM) and self-inhibition at higher concentration (apparent dissociation constant for inhibition, Kb = 9 mM), but the overlap of these two actions only reduces the maximum observable flux by 20%. Increasing ethanol concentration from 0 to 0.9 M causes: a linear increase in the maximum response of the nAcChoR to suberyldicholine from 5 to 80% of the maximum induced by acetylcholine, a moderate increase in Ka, and no change in Kb. Analysis of our results using the sequential two-site binding model revealed that the main action of ethanol on nAcChoR was to increase the fraction of occupied receptors that open. The equilibrium constant describing this effect changed by 8-fold at anesthetic concentrations. Ethanol also decreased the affinity of suberyldicholine for its self-inhibition site by a comparable amount, suggesting that its main action is to stabilize the open state. In addition, ethanol caused a small increase in suberyldicholine's affinity for the agonist site.