The heavy load of individual transport in urban regions induces a demanding request for parking space in conjunction with an ever increasing parking search induced traffic. One possibility to solve these problems is to inform drivers about the parking situation in their destination area. This paper describes a new approach to forecast modelling the occupancy rate, not only of automatically detected parking lots, but also of on-street parking places, taking into account current and future events which influence the general demand for parking space. The model will be the central part of a parking infonnation service developed in Munich which provides the users with information about the current and future availability of parking space. As a further added value for car drivers the information is also intended to be integrated in the route planning of online route-guidance services as the PTA (Personal Travel Assistance). For the covering abstract see ITRD E114174.