Abstract: In order to obtain active peptides with good activity of α-glucosidase inhibition, oyster was used as raw material, and neutral protease and animal proteolytic enzyme were selected for enzymolysis. The inhibition rate of oyster peptidesα-glucosidase was used as index to design the uniform design experiment.Results showed that the temperature of enzymolysis, the amount of enzyme and the concentration of substrate were correlated with the inhibitory rate of α-glucosidase. The three formers were used as inputs to the BP neural network and the latter was used as the output to develop neural network. The process of oyster enzymatic hydrolysis was simulated and the activity of oyster peptides was predicted.The optimal enzymolysis parameters were obtained. The results showed that the maximum inhibition rate of α-glucosidase was 89.22% when the enzymolysis temperature was 55 ℃, the enzymolysis time was 5 h, the enzyme concentration was 600 U/g, the substrate concentration was 0.25 g/m L. Therefore, BP neural network can be used to simulate the non-linear process of oyster enzymolysis, and the α-glucosidase inhibition rate of the hydrolyzate can be well predicted, which is favorable for the industrial production of oyster enzymatic peptides.