Abstract: Side scan and synthetic aperture sonars are widely used imaging systems in the underwater environment. They are relatively cheap and easy to deploy, in comparison with more powerful sensors, like multibeam echosounders. Although side scan and synthetic aperture sonars does not provide seafloor bathymetry directly, their records are finally related to seafloor images. Moreover, the analysis of such images performed by human eye allows creating semispatial impressions of seafloor images obtained from side scan sonar echograms. In the paper, some techniques for 3D seafloor shape reconstruction from side scan and synthetic aperture sonars are presented. They are based on Shape From Shading (SFS) approach, which is one of classical problems in computer vision. The method for reconstruction of 3D seafloor relief using the information from both the currently processed and previous ping is presented. The advantage of the presented methods is their simplicity and the ability to produce the results within sequential, i.e. “one run” processing of side scan sonar image. Another algorithm relies on estimating the altitude gradient of the insonified surface from sonar data, combined with the use of dimension of shadow areas for estimation of the current elevation change. The presented results are promising and also show to some extent how the performance of the proposed algorithm might be improved in further investigation.