In order to evaluate the clinical applicability of a neurophysiological technique, information is required about the effect of technical and subjective factors on the measured parameters, and whether these measures are reliable. The reliability of measurement of a given parameter is inversely related to the variability shown by a set of measurements of that parameter made on the same subject under similar conditions. Reliable measures are fundamental to the clinical utility of any technique. We evaluated the reliability using the intraclass correlation coefficient (ICC), and assessed both single-point and test-retest normal limits for H-reflex parameters of abductor brevis hallucis (ABH). The H reflex from the ABH muscle was recorded in 36 of 43 subjects, but could not be elicited in 7 of 11 subjects who were >60 years of age. The Hmax latency increased as height increased, whereas Hmax amplitude and Hmax/Mmax amplitude ratio decreased as age and height increased. The Hmax latency and Hmax/Mmax amplitude ratio, but not Hmax amplitude, proved to be reliable. The usefulness of the H reflex from ABH in the evaluation of S2 radiculopathies is described.