A real-time monitoring system for grain moisture content in grain storage was developed based on equilibrium moisture model, wireless sensor network and embedded technology. The error and precision which forecast by the equilibrium moisture model of modified Henderson, modified Chung—Pfost and modified Owin were verified in this experiment. The results showed that there was high precision in the modified Chung—Pfost model. Moreover, the prediction result was with higher accuracy when A was 363.06, B was 0 .1804, and C was 26.674. It was indicated that the system was able to realize stable data transmission by testing the transmission quality of sensor nod. This study presented a new method for real-time monitoring rough moisture content in grain storage, which was important significance for ensuring grain safety in storage.