Taking the stock volume of continuous forest inventory in Miyun District of Beijing as research object, and combining auxiliary factors associated with stock volume, the spatial interpolation analysis of the stock volume was carried out by using the ordinary Kriging and Co-Kriging methods, and the results were compared with those of reference 25 in the same study area based on the partial least squares regression method. The results show that based on the auxiliary information, Co-Kriging method is superior to ordinary Kriging and partial least squares regression method, the correlation coefficient between the estimated value and the measured value based on Co-Kriging method was 0.845, the correlation coefficient between the estimated value and the measured value based on ordinary Kriging method was 0.389, and the correlation coefficient between the estimated value and the measured value based on partial least squares regression method was 0.766, respectively. Co-Kriging can significantly improve the prediction accuracy compared with ordinary Kriging, generating the root mean square error decreased by 71%, respectively, and the correlation coefficient between predicted values and measured values increases by 54%. Finally, the spatial distribution map of forest stock volume in Miyun was generated. The research shows that the application of geo-statistical methodology has a good application prospect, and it can provide a feasible method for the estimation of forest stock.