Forty-eight, healthy, adult, volunteer subjects (20 males, 28 females) who ranged in age from 21-46 years were studied during isokinetic testing of the knee joint. The same knee of each subject was tested at speeds of 60, 180, and 240°/sec on both the Biodex Model 8-2000" and the Cybex I 1 Plus' isokinetic dynamometers. The following variables were compared: peak torques, angle of peak torques, peak torque to body weight ratios, and flexion to extension ratios. Analysis of correlation coefficients showed statistically significant (p < 0.05) relationships between several pairs of values. However, results of paired t-tests revealed many statistically significant (p < 0.05) and nonuniform differences. The nonuniform differences between data generated by the two brands of isokinetic dynamometers indicated that clinicians should exercise caution when attempting to extrapolate test data from one machine to another.Isokinetics, a concept introduced in 1967 (36, 59), has gained popularity in a variety of clinical applications. lsokinetic technology is used to strengthen muscle and to assess dysfunction (19,20,29,39,42,43,56,60). Other uses for isokinetic dynamometers include passive stretching, preemployment screening, industrial rehabilitation, determining return to competition status, and research (1 1, 24, 45, 46 0190-601