Amyotrophic lateral sclerosis (ALS) is a degenerative disease of unknown cause, resulting in the progressive loss of voluntary motor activity. Traditional methods of evaluating patients with ALS (neurologic assessment, manual muscle testing and rating scales) involve subjective elements and lack the sensitivity needed to detect small but meaningful changes in deterioration and therapeutic efficacy. This paper presents a recently developed strength measurement instrument, useful for the long-term monitoring of muscle strength deterioration in ALS. In addition, a reproducibility study to assess the intra- and interobserver reliability of maximal voluntary isometric contraction is reported. The strength measurement instrument consists of a strain gauge force transducer, a bridge amplifier and a mechanical structure to counteract movements. A personal computer acquires the strength signal during each voluntary contraction and automatically computes the maximal value. Reproducibility of measurement was assessed in 18 normal subjects tested twice, on successive days, with the same examiner and under the same conditions. The interexaminer reproducibility of measurement was assessed in a group of 13 normal subjects and 10 ALS patients. Each subject of the two groups was independently examined by three examiners. The two reproducibility studies showed a high intraclass correlation coefficient (0.91-0.97) and low SEM and measurement error (3-10%) in all muscle regions tested. The Bland and Altman plots confirmed these results.