A deductive approach to the problem of determining the movement and effects of spilled oil over the Outer Continental Shelf requires that the potential paths of oil be determined first, in order that critical subareas may be defined for later studies. The paths of spilled oil, in turn, depend primarily on the temporal and spatial variability of four factors: the thermohaline structure of the waters, the circulation of the water, the winds, and the distribution of suspended matter. A review of the existent data concerning these factors for the Baltimore Canyon Trough area (a relatively well studied segment of the Continental Shelf) reveals that the movement and dispersal of potential oil spills cannot be reliably predicted. Variations in the thermohaline structure of waters and in the distribution of suspended matter are adequately known; the uncertainty is due to insufficient wind and storm statistics and to the lack of quantitative understanding of the relationship between the nontidal drift and its basic driving mechanisms. Similar inadequacies should be anticipated for other potentially leasable areas of the shelf because an understanding of the movement of spilled oil has not been the underlying aim of most previous studies.