Twenty-eight patients with a clinical diagnosis of carpal tunnel syndrome underwent extensive neurographic examination. The results were compared with those in 86 normal subjects who had no signs or symptoms. The widely used methods of measuring distal motor latency from wrist to thenar and the sensory conduction velocity from the first and third fingers to the wrist gave highly significant differences between the carpal tunnel syndrome group and the control subjects. There was, however, a large overlap of values between the two groups and the discriminative power of the tests was not sufficient to make a diagnosis. Both methods for selectively determining the conduction velocity across the carpal tunnel distinguished clearly between hands with carpal tunnel syndrome and the control group, with no overlap of values. The results support the diagnostic validity of a fractionated study of conduction velocity across the carpal tunnel with surface electrodes, a test that can easily be done by a trained nurse or EMG-technician under the supervision of a clinical neurophysiologist.