Abstract: 【Objective】 The study aims to construct a dynamic model for predicting shelf life of different varieties of walnuts,which provides a theoretical basis for the prediction of storage time and the development of oxidative rancidity control technology of walnut nuts in production. 【Methods】 Wen185,Xinxin 2,Zha 343dried walnut (with shell and without shell) were taken as the test materials, and they were placed in 37℃ incubator respectively. Then the samples were taken regularly to determine acid value(AV), peroxide value(POV) and to make the sensory evaluation. Meanwhile, a chemical reaction kinetic model and Arrhenius equation were employed to construct the model for predicting shelf life of different varieties of walnuts, moreover, and afterwards, the model were applied and verified. 【Results】 The AV and the POV increased and the sensory score decreased in all samples during storage period. The linear fitting function of AV and POV with respect to time was known that the change of AV was consistent with the zero-order reaction kinetic model, the change of POV was consistent with the first-order reaction kinetic model, therefore, the shelf life prediction dynamics model based on AV of the three varieties of walnut nuts(with shell and without shell) was:S L 0=\fracC \quad C 03128.8 e^\frac26009.718.8147; The shelf life prediction dynamics model based on POV of the three varieties of walnut nuts(with shell and without shell) was:S L 1=\frac\ln \fracCC 03128.8 e^\frac26009.718.8147. 【Conclusion】 The error between the predicted value and the measured value was less than 10% of the three varieties of walnut nuts by verified.The model can predict the shelf life of walnut nuts well.