Abstract: Infrared thermocouple temperature (IRT) sensor can estimate crop canopy surface temperature by measuring infrared radiation, on this basis, crop water stress can be monitored. In this paper, the Crop Water Stress Index (CWSI) was applied, which can be calculated by using the crop canopy surface temperature from IRT sensor, to detect the water stress for winter wheat (Triticum aestivum L.) in North China Plain (NCP). The theoretical and empirical models of CWSI were compared in this paper. The results indicate that the empirical model is not the better tool to detect winter wheat water stress in NCP due to its large fluctuation. The relationship between theoretical CWSI and the other parameters of the crop water stress, including soil moisture, leaf water potential, leaf stomata resistances and leaf net photosynthesis, was analyzed. The close correlation between them shows that the theoretical CWSI is a good indicator to detect winter wheat water stress in NCP. This paper gives the main parameters for calculating CWSI for winter wheat in NCP. Considering the feasibility of applying CWSI, the two ways to improve CWSI for monitoring winter wheat water stress and irrigation scheduling were suggested.