A technique is described of measuring the motor end plate jitter in the orbicularis oculi muscle activated by extra-muscular nerve stimulation, standardised for routine use in diagnosis and evaluation of neuromuscular transmission disorders. Among the advantages of the technique are comparatively quick and easy sampling of adequate number of single motor end-plates, convenient control of discharge rate between less than 1 and up to 20 Hz, absence of direct muscle fibre responses, and little discomfort for the patient. The jitter in the orbicularis oculi is significantly lower than that in the extensor digitorum communis muscle. There is a positive correlation between the jitter and the latency of the individual muscle fibre responses, possibly indicating that smaller diameter muscle fibres have slightly lower safety factor of neuromuscular transmission. The data collected in a group of healthy volunteers suggest the following upper normal limits for the mean of absolute consecutive difference of the latency: 30 microseconds for individual motor end plates (one out of 20 values may be higher), and 18 microseconds for a median of 20 motor end plates.