Abstract: Abstract: In this paper, rennet casein was prepared from fresh milk using bacteria rennet as a coagulant and the technology was optimized by a single factor test and response surface methodology. First, temperature, pH value, and an addition volume of enzyme were shown to have significant effects on the yield of rennet casein using the single factor experiment. Subsequently, three significant independent variables were selected and further optimized using the box-behnken design to determine their optimal levels. Via multiple regression analysis on the experimental data using Design-Expert software, the following second-order polynomial equation was obtained. The regression coefficients and the analysis of the variance indicated the high significance of the model. The highest R2 value (0.98) was also in good agreement with the experimental results and theoretical values predicted by the model. From equations derived by differentiation, the optimal values of X1, X2, and X3 in the coded units were found to be 0.306, 0.216, and 0.342, respectively. Correspondingly, we obtained the maximum point of the model, which was 36.53℃ of temperature, 6.216 of pH value, and 0.368 mL of additional volume of enzyme, respectively. The maximum predicted yield of rennet casein was 3.534%. The optimal coagulation parameters in the validated experiment were set as follows: temperature 36.5℃, pH 6.2 and 4 mL 4% CaCl2, 0.37 mL bacteria rennet per 500 mL milk. Under the optimized conditions, the yield of rennet casein reached 3.527%±0.02% and the relative error was 0.19% compared with the predicated value, showing that the response surface method are effective to optimize culture conditions. The contents of water, ash, protein, and fat of rennet casein produced with bacterial rennet and calf rennet were measured. The results showed the contents of water, ash, and protein of rennet casein produced with bacterial rennet were higher than that of calf rennet, but not significantly different (P>0.05), the contents of fat of rennet casein produced with bacterial rennet were lower than that of calf rennet, but not significantly different (P>0.05). The FTIR technique was used to determine and compare the rennet casein produced with bacterial rennet and calf rennet. The spectra of rennet casein produced with bacterial rennet and calf rennet were similar. The results of physical and chemical tests and infrared spectrum scan showed that there was no obvious difference between the two products, suggesting that bacteria rennet from Bacillus amyloliquefaciens could be used as calf rennet alternatives in the production of rennet casein. The research can provide a reference for the full utilization of bacteria rennet.