Sixty-three subjects completed 24 tasks using a menu driven computer program. The menu items appeared in a fixed (static) order during 12 of the tasks. During the other 12 tasks the menu item order changed dynamically such that the most frequently selected items always appeared at the top of the menu. All the subjects tried both dynamic and static menus. The subjects that used adaptive dynamic menus for the first set of tasks were significantly slower than those who used static menus on the first set of tasks. Subjects' performance during the second set of tasks was not affected by menu style. Eight-one percent of the subjects preferred working with static menus to working with dynamic menus.