Abstract: Artificial neural networks are applied for approximation and interpolation of function with multiple variables. Because of concurrent processing of data by neurons, fhis approach can be seen as promising alternate for standard algorithms. From these reasons, the analysis of capabilities for some models of neural networks has been carried out in the purpose for modeling the shape of sea bottom. The feed-forward multi-layer networks with different transfer functions have been tested. These networks have been trained by backpropagation algorithm and its versions with some improvements. Moreover, the gradient optimization technique by Levenberg-Marquardt has been applied. Finally, for determination of the depth in a point of the water area the two-layer network with the hidden layer of the radial neurons has been proposed.