This study proposes an experimental design of biodiesel constrained mixtures to evaluate and optimize the oxidative stability of soybean biodiesel mixed with different amounts of tallow and castor bean biodiesels. To prepare the biodiesel blends, the proportions of each type of biodiesel were varied, producing mixtures that contained 80-86% (v/v) soybean, 5-15% (v/v) tallow, and 1-10% (v/v) castor bean biodiesel. The induction period of pure soybean biodiesel served as the baseline for comparing the stability of the mixtures. The experimental design strategy was capable of producing the best oxidative stability for biodiesel mixtures. The quadratic model obtained in the optimization stage fits the experimental data well. The higher induction period occurred at the following composition: 82.5% (v/v) soybean, 7.5% (v/v) tallow and 10.0% (v/v) castor bean.