The second lumbrical-interosseus distal motor latency (2LI-DML) was compared prospectively in 450 hands. Median nerve function was assessed by standard motor and sensory electrophysiological tests. In a control group of 100 hands the upper limit of normal for the 2LI-DML was 0.5 ms. In all hands studied the correlation coefficients of 2LI-DML were higher with sensory nerve tests than with motor studies. Carpal tunnel syndrome (CTS) was diagnosed clinically in 276 hands, and 174 showed no clinical signs of CTS. The 2LI-DML was prolonged in 269 of the 276 hands, with clinical signs of CTS and normal in 170 of 174 non-CTS hands. Thus the 2LI-DML resulted in a sensitivity of 97.5%. On the other hand, combining the standard tests yielded a sensitivity of 98.5 %. In 31 of 36 additional hands a lumbrical response was recorded, although motor and sensory responses form standard median nerve conduction studies were absent, and the 2LI-DML was substantially prolonged. The 2LI-DML therefore represents a highly sensitive, fast, easy-to-perform, and cost-efficient method to study median nerve function across the wrist and may help to localize the lesion in cases in which standard electrophysiological methods fail.