The purpose of this study was to determine the intrarater and interrater reliability of mean peak torque values for the reciprocal motions of plantarflexion/dorsiflexion (PF/DF) and inversion/eversion (INV/EVER), generated isokinetically at 60 and 1 20°/ sec. Forty-one healthy subjects, ranging in age from 20-75, were tested on three different occasions. The three test sessions were performed at the same time of day and separated by at least 24 hours. Pearson product-moment correlations were used to determine reliability for all ankle motions tested. lntrarater reliability of peak torque values for PF/DF and INV/EVER at both speeds ranged from 0.78-0.94, while the range of interrater coefficients was from 0.82-0.94. The results of this study suggest that when using a well defined clinical protocol, acceptable reliability for both testretest and interrater situations can be obtained for isokinetic peak torque values of the ankle musculature at low and relatively high test speeds.