The knee extension force of 50 patients was investigated using traditional manual muscle testing and hand-held dynamometry. The relationship between manual muscle test word scores and dynamometer force scores was determined using Kendall tau, as was the relationship between manual muscle test percentage scores and dynamometer scores expressed as a percentage of "normal." Percentage scores were also compared to determine if a significant difference existed. Manual muscle test scores and dynamometer test scores were significantly correlated (p less than .001). Percentage manual muscle test and dynamometer test scores were significantly different (p less than .001). These results suggest that the two procedures measure the same variable-strength. Manual muscle test percentage scores of knee extension may, however, overestimate the extent to which a patient is "normal."