An efficient global optimization procedure is presented by using Taguchi's design of experiments (TDE) as a means for undertaking biomechanical studies that rely on experimentations. The proposed TDE is a systematic method of fractional factorial designs for conducting experiments with many independent variables. The approach can provide a step-by-step means for predicting the results of a comparative full factorial design only with a small number of tests. In this study a three-level, four-variable heel-toe running model, and a two-level, seven-variable bicycle example were examined to show the capability and robustness of TDE. In the process of the analysis, the orthogonal array and signal-to-noise ratio analysis of TDE were used to set up the trial conditions and analyze the results. It is shown that in the heel-toe running analyses the TDE successfully predicted the optimum sets of variables with 89% fewer trials than the full factorial design. The reduction in number of trials in the bicycle example is 94%. As a result, the use of TDE analysis to replace a full factorial analysis can considerably reduce the number of trials and still provide a useful outcome in many multifactor biomechanical studies.