A computer-assisted instrument, the isotechnologies B-200, was compared with double goniometer technique for measuring flexion-extension and lateral bending and a compass technique for measuring axial rotation. Test-retest measurements were performed, and the reliability of each method was determined. Generally, low correlations were obtained between two tests performed 1 week apart. The B-200 had poor reproducibility in flexion-extension, particularly when measured separately. Somewhat better results were obtained in lateral bending and rotation. With the goniometers r values close to 0.8 were obtained in flexion, and statistically significantly repeatable data were also obtained in combined flexion-extension and right rotation. Because of the poor repeatability statistical analysis comparing the two instruments was not meaningful. The authors conclude that an instrument such as the B-200 should not be used to measure range of motion. Inexpensive tools allow accurate measurements of, at least, flexion and extension. Small changes should not be taken to indicate true improvement or deterioration. Measurements to a single degree are not meaningful.