A cross-sectional study was performed to assess the utility of vibrotactile thresholds (VTs) obtained before and after a 10-minute period of wrist flexion as a method for detection of carpal tunnel syndrome (CTS) among adult subjects. Subjects with hand discomfort were recruited from patients referred to a university-based electromyography laboratory. Asymptomatic subjects were recruited from among office and technical staff at a professional school. In addition to electrophysiologic evaluation (EP), all subjects were offered VT measurement of the index and small fingers, bilaterally, before and after a 10-minute period of wrist flexion. A total of 144 subjects were recruited, and three hand-condition groups were established: 57 hands had symptoms and EP results compatible with CTS (Group 1), 58 hands had symptoms compatible with CTS and normal EP results (Group 2), and 123 hands had no symptoms and normal EP results (Group 3). Group 1 was considered the "disease-positive" group, and Groups 2 and 3 were both considered "disease-negative" groups. Analyses were performed separately for dominant and nondominant hands, and results were pooled when appropriate. Outcomes of interest were the VTs obtained from the index and small fingers before and after 10 minutes of maximal voluntary wrist flexion as well as variables calculated from them. Significant differences in mean VT were observed between the three hand-condition groups for most of the outcomes evaluated. At any given level of specificity, the sensitivity of vibrometry performed after 10 minutes of wrist flexion was approximately two times that obtained before wrist flexion for detection of electrophysiologically confirmed CTS.(ABSTRACT TRUNCATED AT 250 WORDS)