Surface myoelectric signals are recorded in motor nerve conduction, fatigue and kinesiologic studies using discrete electrodes. Single site recordings have limited means to reduce cross-talk and to enhance timing and quantification of relative muscular activity. These limitations are compounded by the effects of the electrode size. A grid electrode would reduce some of these limitations. However, an optimum grid electrode requires detail examination of the effects of the size of individual electrodes and the interelectrode distance. The purpose of this study is to investigate the temporal and spatial effects of the electrode size on surface motor unit potentials (SMUP). Muscle fiber action potentials and surface electrodes are simulated by computer models. Peak to peak amplitude, the mean frequency of SMUP, and the muscle conduction velocity were calculated as functions of the size of the electrode. The random variations of these parameters due to systematic errors are also simulated and investigated.