An optical path passing through an object-side lens group is bent at a right angle by an optical device so as to be a vertical optical path. A lens group on imaging surface side for forming an image on an imaging surface is disposed on the vertical optical path. The lens group on imaging surface side moves along the vertical optical path, and has a plurality of movable lens groups at least for zooming. The cam plate moves in a direction perpendicular to the vertical optical path within a plane perpendicular to an optical axis of the object-side lens group and not protruding toward an object side compared with the object-side lens group. A plurality of cam slots are formed on the cam plate. The cam slots are fit to a part of the plurality of movable lens groups. In accordance with the movement of the cam plate, the cam slots move the plurality of the movable lens groups such that zooming and focusing are performed.