In a water wet porous medium, the antecedent moisture content controls the entrapment of LNAPL at low capillary numbers. A two-phase retention cell study of entrapment was conducted for LNAPL water systems in a carbonate sand. For two different LNAPL water systems, if was found that a linear expression related the residual LNAPL saturation to the antecedent water saturation. The prediction of trapped LNAPL saturation compares favorably to long-column test data obtained for the same sand. A trapping model was developed, and its predictions were compared to field measurements of trapped LNAPL saturation. Deviations between predicted and measured LNAPL saturations arose when there was a change in the median pore size distribution. In addition, a series of three- phase (air- LNAPL-water) retention cell tests measured the amount of LNAPL that became trapped when the water phase was maintained at a constant saturation during LNAPL withdrawal under a negative pressure head. Under these conditions, the amount of trapped LNAPL was higher than that measured by two-phase tests at the same water saturation.