Abstract: The crop's spatial-temporal pattern is critical to many agricultural studies. In the last 30 years, the rice area in Northeast China increased significantly, and has become one of the major rice producing regions in China. To explore the rice spatial-temporal change characteristics and enrich the crop spatial distribution information acquisition methods, this study combined such multi-source datasets as agricultural statistics data, cropland data, irrigation map and crop suitability, and uses a crop spatial distribution model SPAM (Spatial Production Allocation Model) which has been based on the cross-entropy theory and applied successfully in Brazil and Africa. We further modified the model and applied it to China, and named it SPAM-China. This model mainly has four modules that include a data consistency processing module which makes the multisource data a better spatial match, an agricultural statistical data input module, and a multisource data discrimination module that uses GAMS to optimize the procedure and improve the efficiency of the model, and a result output module. With the SPAM-China model, the study simulates the rice spatial distribution of Northeast China on the pixel scale in the past 30 years (1980-2008) and obtains four rice distribution maps.