KANNUS, P. and K. YASUDA. Value of isokinetic angle-specific torque measurements in normal and injured knees.Med. Sci. Sports Exerc., Vol. 24, No. 3, pp. 292–297, 1992. This study assessed the relationship between the isokinetic peak torque (PT) (speed of movement 60 and 180 degrees · s-1) and the angle-specific torques (ASTs) at 15 and 75 degrees of knee flexion in multiple contractions of the quadriceps and hamstrings in 80 subjects with a grade I knee injury. At every test speed, the Pearson product moment (r) and the Spearman rank (rs) correlation coefficients between the PT and ASTs were highly significant (P< 0.001) for both muscle groups in the uninjured (r= 0.73–0.88, rs= 0.75–0.89) as well as in the injured (r= 0.63–0.91, rs= 0.65–0.91) knees. In both extremities, a great majority of the correlation coefficients exceeded 0.80, which was the predetermined threshold for the coefficients to exceed before the relationship could be considered clinically important. Using regression analysis, both extremities showed completely nonsystematic distribution of the residuals. The variability of the AST measurements was clearly higher than corresponding PT measurements. In conclusion, concerning healthy or grade I injured knees, the predictability of ASTs from PT was good; and, therefore, AST analyses offer little additional information about thigh muscle function than that obtained through the simpler and more commonly used measurement, PT analysis. Because of large variability, the routine use of the AST measurements for grade I knee injuries is not recommended.