When evaluating plants, in particular perennial species, it is common to obtain repeated measures of a given trait from the same individual to evaluate the traits' repeatability in successive harvests. The degree of correlation among these measures defines the coefficient of repeatability, which has been widely utilized in the study of forage traits of interest for breeding. The objective of the present study was to evaluate the repeatability of agronomic traits in Panicum maximum hybrids. Hybrids from three progenies totaling 320 hybrids were evaluated in an incomplete-block design, with consideration of production and morpho-agronomic traits. Of the production traits, total dry matter and leaf dry matter showed the highest repeatability and varied from 0.540 to 0.769, whereas stem dry matter had lower coefficients (0.265-0.632). Among the morpho-agronomic traits, plant height and incidence of Bipolaris maydis had higher coefficients (0.118-0.460). The repeatability values of the agronomic traits were low-to-moderate, and six evaluations were sufficient to provide accuracy in the selection of hybrids regarding total dry matter, leaf dry matter, plant height, and incidence of B. maydis, whereas the other traits require more repeated measures to increase reliability in the prediction of their response.