Experiments were conducted to determine the influence of extraction method, moisture and oil content, soil type, extraction time, sample stirring, method of oil-spiking and drying of samples on oil recovery. Two extraction methods, Dean & Stark and Soxhlet, were tested. Soil moisture showed a highly significant effect on oil recovery. A quadratic regression equation of the Dean & Stark extraction method provided an excellent description of the relationship between oil recovery and soil moisture content, and therefore can be used to quantify oil recovery from soil samples with different moisture levels. The equation predicted a decline in percent oil recovery when soil moisture exceeded 30%. The Dean & Stark method was more efficient in extracting oil, especially at higher oil and moisture contents. Both oil and water recoveries of water saturated samples increased with extended extraction time and sample stirring, but air- and oven-drying of contaminated samples caused oil to volatilize and thus could lead to an underestimation of oil contents. The water content of the soil prior to being spiked with oil did not affect oil recovery efficiency. In a mass balance experiment we were able to account for more than 97% of the total constituents (oil, water and solids) in the samples. The equation we developed provided a more accurate quantification of oil recovery from soil at different moisture levels than was previously available.