Accelerated shelf-life testing and Arrhenius equation were adopted to establish the shelf-life prediction model for lard cookies. With peroxide value as the deterioration index, peroxide value curves of lard cookies during different storage under shelf-life storage condition(25,35,45,55 ℃) was in line with first-order reaction kinetic model. The influence of the number of repetitions, the number of detection points and the time interval on the prediction of the shelf-life prediction model under accelerated storage conditions was studied. The results showed that the number of detection points has the greatest influence on the prediction accuracy, followed by the number of detection repetitions, and the least is the detection interval. The rate constants at different temperatures and the Arrhenius equation were used to establish the self-life prediction model for lard cookies. The shelf-life-prediction model of lard cookies has a good forecasting effect at 25 ℃, with forecasting error -2.74% and has certain application value.