An oilspill risk analysis was conducted to determine the relative environmental hazards of developing oil in different regions of the Beaufort Sea, Alaska, (Proposed Sale 71) Outer Continental Shelf (OCS) lease area. The probability of spill occurrences, likely movement of oil slicks, and locations of resources vulnerable to spilled oil were analyzed. The model predicted movement of the center of spill mass and estimated the times between spill occurrence and contact with various resources, to allow a qualitative assessment of oil characteristics at the time of contact; no direct computation was made of weathering and cleanup. The model also assumed that any oil spilled under ice would remain in place, unchanged, until spring breakup. Ice movements, or travel of oil under ice, if occurring, would affect the results in a manner not directly predictable at this time. The combined results of spill occurrence and spill movement predictions yielded estimates of the overall risks associated with development of the proposed lease area. Assuming that oil exists in the lease area (a 99.3-percent chance) it is estimated that the leasing of the tracts proposed for OCS Sale 71 will result in an expected 9.2 oilspills (of 1,000 barrels or larger) over the lease lifetime of 25 years. This estimate is based on historic oilspill accident data for platforms and pipelines on the U.S. OCS (Gulf of Mexico and California). The estimated probability that land will be contacted by one or more oilspills (of 1,000 barrels or larger) that have been at sea less than 30 days (not counting any time trapped under ice) is greater than 99.5 percent. If oilspill accident data for Prudhoe Bay, Alaska, is used in the analysis, it is estimated that 5.6 oilspills (1,000 barrels or larger) will occur over the lease lifetime. The estimated probability that one or more oilspills (1,000 barrels or larger)will occur and contact land is99 percent. The results of a recent experimental cleanup operation for oilspills under ice in Canadian Arctic waters showed a substantial degree of success; this should be considered in evaluating impacts of spills predicted by this model. Oilspill occurrence probabilities are high primarily because of the large amount of oil believed to be present in the area.