ABSTRACT Absolute muscle strength or adjusted by body mass index (BMI) are useful to identify sarcopenia. However, these values are not accurate for older adults with extreme body sizes because the nonlinear relationship between strength and body size. The purpose was to determine cut-off points for identify sarcopenia in older adults using allometric coefficients to normalise handgrip strength (HGS) and 30-sec chair stand test (30-s CST) by body size. Allometric exponents were proposed with log-linear models for body-size variables (body mass, height and BMI). The remotion of body-size effect on muscle strength with allometric normalisation was tested by partial correlation. Cut-off points for low muscle strength were established by ROC curve and Youden index considering functional limitation (six-minute walk test<400m). Allometric exponents provided for body-size variables range from -0.01 to 2.28 (HGS) and -0.27 to 0.21 (30-s CST). The effect of body size on muscle strength was removed with allometric normalisation (r<0.30). Cut-off points accuracy was always adequate (AUC≥0.78; p<0.001). In conclusion, cut-off points of HGS and 30-s CST allometrically normalised were proposed to identify sarcopenia in Portuguese older adults and allometry maintained adequate the accuracy (AUC>70%). Allometry removed influence of body size on the expression of HGS and 30-s CST and permits evaluate muscle strength regardless of body-size.