Linear and nonlinear optimal designs have been used abundantly to predict the forces exerted by individual muscles for everyday movements such as walking. Individual muscle force predictions for athletic movements, those involving large ranges of motion and fast velocities of muscle contractions, are almost nonexistent. The purpose of this paper is to illustrate some of the design characteristics that must be considered for predicting individual muscle forces in athletic movements. To do this, the load sharing between two muscles, derived from nonlinear optimal designs, is considered in two ways: (a) in hypothetical experiments of muscle contractions, and (b) in real experiments of knee extension movements performed by one subject. The results suggested that additional design considerations must be made when predicting forces in athletic movements compared to everyday movements.