The method of successive approximations in the strategy space is used to solve a nonlinear inverse problem. The method is based on gradient analysis; approximation to the optimal strategy is achieved by successive steps. Well-known concepts and approaches of dynamic programming are used to substantiate the method.