Selecting an obscured object of a computer-generated model consisting of a plurality of objects displayed on a computer screen includes receiving input data interpreted as an instruction to change a visibility characteristic of a first one of the plurality of objects (the first one of the objects obscuring another of the objects, and the first object being determined based on the location of an on-screen cursor), and modifying a visibility characteristic of the first object to cause another one of the objects to become discernable and selectable.