Abstract: The high accuracy, continual distribution and natural rules of light-temperature crop productive potentility and climate crop productive potentiality are analyzed based on the spatial interpolation method, which improve the sciefication, accuracy and practicality of agricultural land classification. Compared the common spatial interpolation models and based on the classification data of agricultural land in Henan Province, the initial interpolation results of Kriging are better than those of other models. The initial interpolation value should be modified in mountainous area and the area of absent data because of the limitation of model. The method of modifying the initial interpolation value of light-temperature and climate crop productive potentiality were put forward.