You can have a query either search for multiple years using OR , or search for a year in a string of years by using LIKE.
However, the best solution is, as NeoGeo says, to put your year column into a seperate table. I would go further and say, Normalise your data properly. I would guess that it is possible for an item to fit more than 1 make, or more than 1 model, as well as more than 1 year. You should generalise your table designs to allow for this. It will impress the client when you point out that you have done so.