Optimal cutoff frequency of bi-directional zero phase-lag Butterworth digital filter, determined through the technique of residual analysis, was evaluated using the methodology of computer simulation. Squat jumping (SQJ) and counter movement jumping (CMJ) motions were analyzed. A three-dimensional neuromusculoskeletal computer simulation model of the human body was constructed. Realistic SQJ and CMJ motions were generated through forward dynamic computer simulation employing an extensive procedure of numerical optimization. Residual analysis was performed on each coordinate in order to obtain the optimal cutoff frequency. Effects of artificially introduced experimental noise (Gaussian noise) were evaluated. It was found that the optimal cutoff frequency was underestimated when the experimental noise was imposed on the kinematic data. In other words, the possibility of an information loss was suggested as a result of digital filtering with the cutoff frequency determined through the residual analysis. It was suggested that the optimal cutoff frequency obtained as a result of residual analysis should be compared with the residual - cutoff frequency characteristics obtained through analyzing noise-free kinematics.