This study investigates vessel oil spill differentials for transfer and vessel-accident spills for the post Oil Pollution Act of 1990 (OPA-90) period. In-water and out-of-water transfer/vessel-accident oil spill equations are estimated, utilizing tobit regression analysis and data of individual vessel oil spills (of oil-cargo and non-oil-cargo vessels) investigated by the US Coast Guard for the 1991-1995 period. In the data, 47.5% (94.1%) of in-water (out-of-water) gallons of oil spilled were transfer spills; non-oil-cargo vessels accounted for 44.2% of all gallons spilled. The estimation results suggest that out-of-water transfer spills are larger in size than vessel-accident spills, but are similar in size for in-water spills. Transfer/vessel-accident spill differentials exist (do not exist) among determinants of in-water (out-of-water) spills. The policy implication of the results is that a vessel oil spill-reduction differentiation regulatory regime is needed that differentiates between transfer and vessel-accident spills and in-water and out-of-water spills for reducing vessel oil spills.