Improving transportation infrastructure raises several questions requiring complex analysis. Creating the mathematic model for transportation reconstruction needed developing a quick, dynamic programming process which considers the costs of single reconstruction phases and the direct or indirect profit emerging after completing reconstruction works. In the study we examined the application possibilities of the highly prestigious Bellman-theorem. We found that the combinatorial method for optimal distribution of investments is not very efficient in the current issue, although it applies the sequential method and the Bellman-theorem in a remarkable way. The method proposed in the study applies exponential partial profit-functions. The primary advantage is that it solves the problem of the complicated nonlinear dynamic programming on the basis of direct relations, without numeric analysis. Consequently, results are supplied quickly and the method allows solving enormous problems without any constraints. © 2007 IEEE.