Miniature endplate currents (MEPCs) recorded from mouse diaphragms with a point voltage clamp, without inhibition of acetylcholinesterase (AChE) and in the absence of any drug, showed in their decay phase consistent deviations from an exponential time course, consisting of (a) "curvature," a progressive increase of decay rate during most of the decay phase, followed by (b) "late" tails. Both phenomena persisted when MEPCs (and channel lifetime) were prolonged by ethanol. Curvature was increased by muscle fiber depolarization and decreased by hyperpolarization. Receptor blockade by (+)-tubocurarine, alpha-bungarotoxin, hexamethonium, or myasthenic IgG accelerated the decay of the main part of MEPCs and eliminated curvature; the time constant of MEPCs became close to the channel time constant. We conclude that curvature arises from repeated action of ACh with cooperativity in ACh-receptor interaction; the voltage sensitivity of curvature follows from the voltage sensitivity of channel closing. Ethanol, in addition to its effect to prolong channel lifetime, enhances the tendency of ACh to act more than once to open channels before being lost to the system. Analysis of the rising phase of the MEPC, in terms of driving functions, also indicated that ethanol promotes channel opening by ACh; this action can account for a substantial increase of MEPC height by ethanol when MEPCs are made small by receptor blockade. Driving functions were also voltage sensitive, in a manner indicating acceleration of channel opening, but reduction of channel conductance, with hyperpolarization. Poisoning or inhibition of AChE prolonged MEPCs without altering the duration of ionic channels. Since ethanol caused further prolongation of MEPCs after poisoning of AChE, with little change in MEPC height, we conclude that the extension of mean channel lifetime by ethanol is accompanied by a similar extension of ACh binding to receptors. After poisoning of AChE, MEPCs became very variable in time course and the decay rate (tau-1) was correlated with MEPC height with a slope of log tau vs. log height of 0.77 for MEPCs of greater than 60% mean size. This slope is larger than expected from cooperativity in ACh-receptor interaction. Correlation of tau and height of MEPCs also exists when AChE is intact; the slope of log tau vs. log height was 0.12 with or without prolongation of MEPCs by ethanol.