Abstract: Multibeam sonar systems can provide hydrographic quality depth data as well as high-resolution seafloor sonar images. Using the seafloor-backscattered data from each beam and with automatic classification, seabed sediments distribution maps can be obtained directly. In this paper, the self-organizing map (SOM) neural network is used in acoustic seafloor classification from multibeam sonar data. This method can rapidly identify all kinds of seafloor types such as mud, sand, gravel and rock in the experimental surveying areas. Compared with the traditional geologic grab method, the experiment indicates that the SOM method is feasible and valid.