An image forming apparatus includes a plurality of container groups, and a container housing. The container groups include a plurality of toner containers containing toner of different colors arranged in a left-and-right directions. The container housing houses the container groups in arranged an upper-and-lower direction. The toner containers of the container groups arranged in the upper-and-lower direction are shifted in the left-and-right direction. The container housing includes a lower rail, an upper rail, and an opening frame. The lower rail houses lower halves of the toner containers of the container group. Then upper rail houses upper halves of the toner containers of the container group and is connectable to the lower rail. The opening frame has openings to and from which the toner containers are attached and detached. A container rail assembly in which the lower rail and the upper rail are connected is assembled to the opening frame.