Abstract: To improve the reliability of remote sensing of forest stocking equation and solve the correlation problem among independent variables, this study intended to use the clustering analysis and ridge estimation model to estimate the forest stock in Miyun County. Remote sensing and topographic factors that having an impact on forest stocks were selected as classification characterization factors for cluster analysis. According to the clustering results, the modeling samples were selected, and the forest estimation volume was estimated by the ridge estimation model, and the applicability evaluation and accuracy verification were carried out. Three kinds of classification results were obtained by aggregating the sum of squared deviations, the model samples were extracted by weights, and the models were verified by 30 independent reserved samples. The results showed that the R2 of the measured value and the estimated value of the reserved sample was 0.5311, the root mean square error was 1.4553, the relative deviation was 8.9%, the measured value was 90.942 m3, and the estimated value was 82.842 m3. The applicability of the model was general, the estimation accuracy reached 91.1%, and the overall estimation accuracy was high.