The records of 138 patients (222 hands) who had been evaluated by preoperative nerve conduction study (NCS) and consecutive self-administered questionnaires were analyzed to determine diagnostic relations in carpal tunnel syndrome. Hands were categorized into six grades (extreme, severe, moderate, mild, minimal, and negative) according to the three modalities: the NCS data of the median sensory nerve, Symptom Severity Scale score, and Functional Status Scale score. The authors evaluated the correlation, independence, homogeneity, and association among parameters, and found the following: (1) the NCS data correlated more significantly (p < 0.05) and was associated more linearly with Symptom Severity Scale score than the Functional Status Scale score; (2) the choice of modality exerted an influence on grading severities, thus grades varied according to the modalities; and (3) the distribution of patients was not homogenous; there were more extremes in NCS, and more severe scores on the Symptom Severity Scale. Patients rated as one homogeneous severity by NCS data may be interpreted into different severities (from negative to extreme) by self-administered questionnaires and vice versa.