Abstract Bridge‐water clearance controls safe vessel passage and reflects local water‐level changes, but direct measurements are unavailable at many bridges. We present BridgeSAR, a geometry‐guided method that estimates clearance from multipath scattering stripes in synthetic aperture radar (SAR) amplitude images. BridgeSAR uses OpenStreetMap bridge geometry to guide extraction of single‐, double‐, and triple‐bounce stripes, accumulates SAR amplitude energy normal to the bridge axis, and converts stripe spacing to clearance using viewing geometry and a fixed structural‐depth term. We apply the method to Sentinel‐1 images from 2016 to 2025 over five U.S. bridges and compare the results with air‐gap sensors or water‐level‐derived reference clearances. The single‐to‐triple‐bounce separation provides the most stable observable, reproducing dominant tidal and seasonal clearance variations at most sites, with best root mean square errors of 0.39–0.61 m. BridgeSAR performance is influenced by viewing geometry, stripe signal‐to‐noise ratio, and wind‐driven roughening of the water surface.