The use of powered wheelchairs with high navigational intelligence is one of the great steps towards the integration of severely physically disabled and mentally handicapped people. Driving a wheelchair is a difficult task even for a normal person and it becomes much more difficult for people with disabilities. This paper describes the use of the case-based reasoning (cbr) paradigm in this field. The objective is to discharge the user from control constraints by assisting him in the reproduction of routes frequently employed. The assistance task is based on two important steps. The first one consists in recognizing the followed route based on a similarity computation with one already carried out. In the second, future directions of the wheel- chair are predicted in order to reduce the number of com- mands which the user has to define. An automatic learning of routes is proposed allowing the adaptation of the wheel- chair behavior in different environments.