Nowadays, larger and higher resolution screens supporting multiple windows are widely available. They are often used to traverse hypertext databases for fact retrieval, education or casual browsing. This paper describes a study comparing two methods of window control strategies for article placement in four, equal sized and tiled windows (By tiled we mean non-overlapping, non-movable and non-resizable windows, which together cover the entire screen.) The first placement strategy - Automatic Panning (A P) - was controlled by the system software and the second placement strategy - User Controlled Placement (UCP) - trasferred control to the users. The task required subjects to answer questions about the content of a hypertext. Statistical analysis shows that subjective preference was significantly higher and amount of backtracking was significantly lower for the user controlled placement. Speed differences did not prove to be significant, although there was consistent tendency favoring the automatic pa nning strategy.