Abstract: The accuracy of traditional cloud service platform is low when processing large amounts of data. Therefore, the intelligent agriculture cloud service platform based on BP neural network is designed. In terms of overall architecture design, PHP, JavaScript and other technologies are combined to divide the overall architecture into three different levels to realize the dynamic interaction of the internal levels of the platform. In the aspect of system software design, agricultural information acquisition module is designed based on D-S evidence theory algorithm. BP neural network model was established to find the optimal value of weight value and fixed threshold value, and to determine the accuracy of model fitting state. B/S and C/A/S structure are adopted to design media library management module, and IP network platform is used to realize data transmission and backup. The test results show that the designed cloud service platform, when the number of information increases to 6000, the average accuracy of the system is 95.17%,which can achieve the expected effect.