1 Subcutaneous injection in mice of a single dose of an organophosphorous anticholinesterase, ecothiopate (0.5 gmol kg-'), produced increased variability in the latency (jitter) of indirectly-elicited action potentials in diaphragm muscles 5 days after treatment, but there was no effect on the variability of latencies of endplate potentials. This study was designed to elucidate the mechanism(s) of the increase in action potential jitter. 2 Action potentials evoked directly by electrical stimulation at one end of muscle fibres and recording near the other end had less jitter than indirectly-evoked action potentials and ecothiopate had no effect on directly-evoked action potentials. 3 In preparations with uncut fibres, pretreatment with ecothiopate reduced by about 20% both muscle fibre input resistance and the amplitude of spontaneous miniature endplate potentials. Ecothiopate had no effect on muscle fibre resting membrane potential or on the threshold potential for excitation. 4 In untreated preparations, indirectly-evoked action potentials recorded at the endplate had similar jitter to action potentials recorded at the tendon when latencies were measured at 10% of peak amplitude. However, when latencies were measured at peak, there was greater jitter of action potentials at the endplate. Ecothiopate increased jitter of action potentials recorded at the endplate at 10% of peak but did not significantly increase jitter of action potentials recorded at the endplate when measured at the peak. 5 In cut-fibre preparations, the first endplate potential of trains was significantly increased after ecothiopate but there was no effect of ecothiopate on the amplitude of plateau endplate potentials later in the train. Analysis of plateau endplate potentials showed that 5 days after administration, ecothiopate produced an increase in the variance of endplate potential amplitudes and changes in the binomial parameters n and p. 6 It was concluded that the increased jitter produced by ecothiopate is not a generalized effect on the plasma membrane and that none of the above observations could explain the increased jitter. The possibility is discussed that increased jitter is produced by variability in times to threshold of endplate potentials and/or by variability in the locus of generation of the action potential in the perijunctional area.