Abstract: Perilla frutescens（perilla）was used as the test materials and stored at room temperature, 2 ℃,4 ℃ and 6 ℃.The effects of different temperatures on quality indicators during storage were studied, and the dynamic model and Arrhenius equation were used to build shelf-lifeprediction model based on characteristic indexes, then the constructed model was validated and analyzed, in order to provide reference for postharvest storage and preservation of perilla for fresh-eating.The results showed that there were significant differences in the postharvest quality of perilla for fresh-eating under different storage temperature conditions（P＜0.05）.Stored under low temperature conditionscan significantly improved the postharvest quality（P＜0.05）.Perilla for fresh-eating was ideal stored under 2 ℃ according to advantages and disadvantages of quality indicators.shelf-life prediction model was constructed based on characteristic indicators of weight loss rate and soluble sugar.With verification and analysis, it concluded that the average relative error between the predicted and measured values was less than 5%.The model had high fitting accuracy and could be used to predict the shelf life of perilla for fresh-eating during stored within the range of 2-25 ℃.