[Objective] The study aimed to establish a model to predict heavy metal pollution in the Xiangjiang River(from Zhuzhou to Changsha) on the basis of Matlab.[Method] According to the data of heavy metal content in the Xiangjiang River(from Zhuzhou to Changsha),we established a non-linear regression model based on Matlab to forecast the content of Cd,Pb,Cu,Zn,As and Cr in the Xiangjiang River.[Result] Verification showed that the prediction models had a high precision,and the spatial variation of the predicted heavy metal content was basically consistent with the actual conditions,which indicated that these models could forecast the spatial variation of heavy metal content in the Xiangjiang River(from Zhuzhou to Changsha) well.[Conclusion] The research could provide theoretical bases for controlling the heavy metal pollution in the Xiangjiang River.