Abstract: Because of the uncertainty of grain drying system, it is hard to realize real-time control using the traditional control method which adjusts discharge rate and ventilating condition according to output grain moisture content. In this paper, a kind of self-adaptive control system for grain drier was developed based on analytical theory of grain deep-bed drying process and precise detection technology of high-moisture-content grain. It can change the system automatically to ensure the optimization of working condition according to real-time input-grain-moisture-content and output status. The control system was proved to be reliable, stable, accurate and practical in production application. For great moisture content differences between kernels, this device can accurately detect the moisture content of each kernel and moisture distribution with the dynamic temperature variation between –30℃ and +40℃ and moisture content variation between 10%wb and 35%wb. And the error of output moisture content is less than ±0.5%wb. So this adaptive control system for grain dryer is reliable and practical for industrial application.