The effect of body size on the output of the tests of directly assessed muscle power has been mainly ignored in the applied research. We hypothesized that the muscle power output would reveal a positive relationship with body size closely in line with the theoretical predictions, as well as that the strength of the studied relationship could justify standardized normalization procedures applied in routine testing. Male physical education students (n = 111) were evaluated in 10 standard tests of direct assessment of muscle power output. A standard allometric relationship P = A SB was applied to assess the relationship between the tested power output P and selected indices of body size S ( B = allometric exponent; A = parameter). The correlation coefficients obtained between the tested power output and body size ranged within 0.21 - 0.56 for body mass and 0.10 - 0.49 for body height. The mean (SD) values of the allometric exponents B obtained with respect to body mass and body height were 0.55 (0.15) and 1.15 (0.62), respectively. This relationship proved to be strong enough to merit applying a standard normalization. The normalization method should be based on body mass as a preferred index of body size, while the allometric exponent could correspond to the theoretically predicted one (i.e., B = 0.67).