This study presents the investigations on the application of genetic algorithms to solve multimodal shortest path problem. To evaluate the efficiency and robustness of proposed method, the algorithm was carried out over 250 randomly selected pairs of origin and destination points with different distances and number of nodes. It was assumed that three modes of walking, bus and subway are used to travel between points. The classification of results in three main classes of monomodal, bimodal and multimodal paths shows that more than 65% of paths are multimodal. These results show the robustness of proposed model. It is also concluded that these experimental outcomes validate the effectiveness of evolutionary methodology to solve the problem.