The National Exposure Research Laboratory of the U.S. Environmental Protection Agency conducts research to measure exposure of individuals to chemical pollutants through the diet. In support of this research, methods are being evaluated for the determination of pesticides in dietary composite samples. In the present study, Soxhlet, blender, microwave-assisted, pressurized fluid, and supercritical fluid extraction methods were compared for the determination of incurred and added pesticides in 4 dietary composites, which varied in fat and water content. Incurred pesticides were chlorothalonil, chlorpyrifos, DDE, dicloran, dieldrin, endosulfan I, malathion, cis- and trans-permethrin, and trifluralin. Added pesticides were alpha- and gamma-chlordane, hexachlorobenzene, and fonofos. Concentrations of the individual pesticides were between 0.2 and 20 ng/g composite. All 5 methods tested could extract pesticides from dietary composites. Most incurred pesticides were recovered from the dietary composites within the range of 59-140% of expected values. Recoveries of added pesticides were between 60 and 130%. Microwave-assisted extraction led to significantly higher concentrations of 7 pesticides. Blender extraction yielded significantly higher concentrations of chlorothalonil and fonofos. Water content was a significant factor in the recovery of chlorothalonil, and fat content was a significant factor in the recovery of fonofos. In designing an exposure study, the selection of the extraction method would be determined by number of samples to be extracted, analyte stability, and cost.