This study investigates the capabilities of the Argentinean satellite SAOCOM, developed by the National Commission for Space Activities (CONAE), to classify native forests in various regions of Argentina using its L-band synthetic aperture radar (SAR). SAOCOM offers high spatial resolution (10 to 100 metres) and multiple acquisition modes, allowing detailed observation of diverse terrestrial ecosystems. Our research focuses on the processing and calibration of intensity and phase data generating bands to classify land cover into Forest/Non Forest. A speckle filter and polarimetric decompositions are employed to improve the performance of the Random Forest classifier bands. Key methodologies include the application of Pauli, Freeman-Durden and Cloude-Pottier decompositions, which allow discriminating the backscattering mechanisms characteristic of different cover types. Both Intensity-derived bands and polarimetric products are evaluated in forest classification (band performance, accuracy, and root mean square error) and compared by cross-tabulation with other tests performed in the same areas. As outstanding results we find that the classification model fits best for intensity and Cloude-Pottier data in arid and semi-arid forests. In the performance of the bands, cross-polarization bands and bands representing the volume backscattering effects of the polarimetric bands stand out. In cross-comparisons with national and international native forest mappings, the classifications using intensity bands and Cloude-Pottier bands stand out, especially for dry forests.