An Artificial neural network (ANN) is a mathematical model that imitates biological neural network properties, and thus approximates operations of the human brain. Neural networks application help to solve difficult problems that the human brain or other computational techniques can’t. Neural networks are being widely used to real world problems; one of them is transportation systems. Nowadays the biggest problem for many cities is a poorly organized traffic management system, traffic flows rapidly increase in last years and cities are often not ready for this. ANN application can give a possibility to improve, optimise the transportation system. ANN’s have already been successfully applied for different prediction, forecasting, classification, detection and estimation tasks. In the paper, the neural networks based approaches (Multilayer, The Kohonen maps, Radial Basis function) for solving different transportation planning problems such as traffic flow, travel time and passenger demand forecasting, sign and moving object on the road recognition, incident detection are described