A special electromyography (EMG) method, scanning EMG, was introduced by Stålberg and Antoni in 1980 to study the electrophysiological cross sections and sizes of motor units. Scanning EMG gives a new approach for the evaluation of the electrical properties of motor units, providing new data on the normal anatomical distribution of muscle fibers and its changes in different pathologies of the muscle. The description of scanning EMG recordings required the introduction of new parameters (lengths of motor unit cross sections, fractions of motor units, and silent areas), in addition to those used with conventional EMG recordings, and the traditional parameters (duration, amplitude, etc.) acquired new and more accurate explanations. Normal scanning EMG recordings are available for biceps brachii, anterior tibial, and masseter muscles. The findings in normal muscles agree with the nonrandom distribution of muscle fibers in motor units and confirm the suggestion that muscle fibers within motor units tend to be arranged in clusters. In muscular dystrophies, the sizes of motor unit territories do not differ significantly from the normal values. However, the configuration of motor units changes considerably. Abrupt changes in amplitude and duration, segments of short and long duration, increased numbers of fractions, and silent areas have been revealed, showing that dystrophic motor units are definitely fragmented. Scanning EMG supports the assumption that there is clustering of muscle fibers within the dystrophic motor unit, with local grouping of muscle fibers. In neurogenic lesions, the length of motor units is normal or only slightly increased. Reinnervated motor units are restricted to the fascicles in which they are originally found. Reinnervation does not result in an increase in the number of fractions, but the amplitude of the potentials, the length of polyphasic sections, and the duration increase. The increase in the number and length of polyphasic sections can differentiate normal motor units from abnormal ones. However, other features (amplitude, duration, number of fractions, and presence of silent areas) are also necessary to distinguish neurogenic processes from myogenic ones.