In this paper, two statistical methods were applied to optimize medium components to improve the production of the milk-clotting enzyme by Bacillus amyloliquefaciens D4. First, wheat bran juice, skim milk powder, and $Na_2HPO_4$ were shown to have significant effects on D4 enzyme production using the Plackett-Burman experimental design. Subsequently, an optimal medium was obtained using the Box-Behnken method, which consisted of 3.31 g/l of skim milk powder, 5.0 g/l of sucrose, 0.1 g/l of $FeSO_4{cdot}7H_2O$, 0.1 g/l of $MgSO_4{cdot}7H_2O$, 0.1 g/l of $MnSO_4{cdot}2H_2O$, 0.1 g/l of $ZnSO_4{cdot}7H_2O$, 1.52 g/l of $Na_2HPO_4$, and 172.45 g/l of wheat bran juice. With this optimal medium, the milk-clotting enzyme production was remarkably enhanced. The milk-clotting enzyme activity reached 3,326.7 SU/ml after incubation of 48 h, which was 1.76-fold higher than that of the basic medium, showing that the Plackett-Burman design and Box-Behnken response surface method are effective to optimize medium components, and B. amyloliquefaciens D4 possessed a high rennet-producing capacity in the optimal medium.