The paper presents a heuristic solution technique for an extended vehicle scheduling problem. The problem is decomposed into five main decision areas: the route assignment, the depot assignment, the vehicle assignment, the delivery period, and the detailed route design. Each part is solved separately and then connected by an iterative procedure. The modular structure makes it possible to work with stochastic demands with the additional constraints of fixed depot and fixed vehicle assignment for each customer. In addition, the method includes the aspects of multiple use of vehicles and the possibility of postponing some of the deliveries. The method should, therefore, be able to handle many different long-term, as well as short-term, distribution planning problems.