We congratulate Dr. Stetson and his colleagues [Stetson et al., 19931 for their decision to use an objective measure of disease [nerve conduction studies (NCS)] in their studies of the relationship between work and carpal tunnel syndrome (CTS). Many of their findings are identical or nearly identical to our own (Table ). However, we find their study to be deficient in several areas. They use a nonrepresentative control group. They excluded from consideration any symptomatic administrative and professional workers, which gives an inaccurate picture regarding the relationship between work and NCS abnormalities. The control standard deviations that Stetson et al. give for the sensory latencies (0.2 ms) are much less than the standard deviations reported by us (0.4 ms) and by others for their control groups [Jablecki et al., 19931. The difference in sensory amplitudes is likely due to gender: females in every group we have studied have higher sensory amplitudes than males (30-60% higher). The authors measured only one hand, the dominant hand, and they failed to control for many confounding variables, such as body mass index, wrist dimensions, and lack of aerobic avocational activities, which have been shown to be more important than work for predicting which subjects and which wrists have NCS abnormalities and CTS [