Estimation of structural forest attributes, such as volume, basal area, and tree density using a combination of remote sensing and field data, is currently considered a favored option compared to only using field survey data. In a comparative study, multiple linear regression (MLR) and classification and regression tree (CART) models were used to estimate volume, basal area, and tree density using advanced space-borne thermal emission and reflection radiometer (ASTER) and satellite poure I’observation de la terre (SPOT)-high resolution grounding (HRG) imagery in the Darabkola forests, located at the Hyrcanian region of Iran. Results showed that the CART model using SPOT-HRG data achieved the best overall performance for all three forest structural attributes, with adjusted R 2 =0.746 and RMSE=67.9 m 3 ha −1 for volume, adjusted R 2 =0.771 and RMSE=3.94 m 2 ha −1 for basal area, and adjusted R 2 =0.871 and RMSE=34.71 nha −1 for tree density. In general, the CART model, using both ASTER and SPOT-HRG data, produced better estimates of forest attributes compared to the MLR model. In addition, results showed that forest attribute estimations using SPOT-HRG were better than those obtained from ASTER data.