Median-ulnar comparative studies (MUCSs) play an important role in the electrodiagnosis of carpal tunnel syndrome (CTS), but diagnostic criteria in patients with underlying axonal polyneuropathy have not been established. We prospectively evaluated 28 patients with axonal polyneuropathy (48 hands), blindly dichotomized into CTS cases and controls, using the lumbrical-interosseous motor (LIM), mixed palmar sensory (MPS), and digit 4 sensory (D4S) MUCS. LIM, MPS, and D4S MUCS latency differences could be calculated in 98%, 69%, and 48% of hands. Based on the maximum combination of sensitivity and specificity, diagnostic cutoffs were > or = 0.8, > or = 0.5, and > or = 0.4 ms for the LIM, MPS, and D4S studies, respectively. The LIM study was the most accurate test and the only MUCS that yielded a positive likelihood ratio greater than 10 (16.25 for a cut-off value of 1 ms) and a negative likelihood ratio less than 0.1 (0.09 for a cut-off value of 0.4 ms), resulting in meaningful differences in pretest and posttest probabilities of CTS in patients with axonal polyneuropathy. These novel MUCS electrodiagnostic criteria should facilitate more accurate electrodiagnostic confirmation of CTS in this population.