Over the past decade, synthetic aperture radar (SAR) sensors have demonstrated their ability to provide accurate wind speed and direction measurements in the ocean. Although SAR capability to image the sea surface at a suitable spatial resolution, SAR wind field retrieval in marine coastal areas still remains an open issue. The aim of this work is to assess the potential and limits of the Bayesian approach to the C-band NRCS SAR inversion problem for the retrieval of wind fields in marine coastal areas. A dataset of 139 SAR images gathered by the European ERS-2 and ENVISAT satellites during 2007 was analyzed to estimate the sea surface wind vectors over three selected coastal sites in the Mediterranean Sea, offshore Puglia region (Italy). The sites were instrumented to monitor marine and environmental parameters, including wind intensity and direction for further comparison with SAR wind estimations. Results discussed in this study show that the statistical approach in the retrieval of SAR wind speed can be applied in coastal areas with performances comparable with those obtained off-shore. In contrast, the SAR inversion procedure was not able to retrieve wind direction with the same accuracy.