In this study we tested and evaluated an algorithm for generating binary snow maps using Sentinel-1 SAR data on various land cover types in midlatitude lowlands. The SAR data was acquired over the Šventoji river basin in Lithuania and included five cold seasons from 2014 to 2021. Snow classification was done by thresholding SAR backscattering ratio between the image with snow and the reference image based on the snow-free SAR observations in autumn. We used ground range dual-polarized Sentinel-1 data and calculated the weighted VV and VH polarization ratio to determine land cover type-specific snow classification thresholds. The validation of snow maps derived from SAR data was done using Sentinel-2 images. Validation results showed that accuracy of SAR based snow maps over grasslands and arable land was within a 0.85–0.92 range, while in the coniferous forests it was 0.49. Results suggest that the most important factors influencing the accuracy of snow detection using Sentinel-1 SAR data are snow depth and dense forest vegetation. The study contributes to the evolving SAR-based algorithms to determine snow cover characteristics and demonstrates the capability of the Sentinel-1 mission for snow monitoring in various landscapes.