N. Walker and J. B. Smelcer (1990) found that selection times for menus could be reduced by placing the menus adjacent to the edge of the screen; this creates a border between the menu and the edge of the screen that the mouse cursor cannot penetrate. Based on this finding, they proposed that selection times for graphical user interface targets could be reduced by employing these impenetrable borders. Four studies tested this prediction with a Web browser's back button and scroll bar. Results demonstrated that targets employing impenetrable borders were always selected faster than were targets placed 1 pixel from the edge of the screen, which supports Walker and Smelcer's prediction. However, within the constraints of the current studies, this speed advantage asymptotes at approximately 283 ms for target heights of 2.00 cm and target distances of 3.50 cm. In addition, these findings generalized across most angles of approach. Actual or potential applications of this research include target placement decisions in the design or modification of graphical user interfaces.