A method for the determination of vitamin D-2 and vitamin D3 in fortified milk powders and infant and adult nutritional formulas is described. Samples are saponified at high temperature and lipid-soluble components are extracted into isooctane. A portion of the isooctane layer is transferred and washed, and an aliquot of 4-phenyl-1,2,4-triazoline-3,5-dione is added to derivatize the vitamin D to form a high-molecular mass, easily ionizable adduct. The vitamin D adduct is then re-extracted into a small volume of acetonitrile and analyzed by RPLC. Detection is by tandem MS, using multiple reaction monitoring. Stable isotope labeled vitamin D-2 and vitamin D-3 internal standards are used for quantitation to correct for losses in extraction and any variation in derivatization and ionization efficiencies. A single-laboratory validation of the method using AOAC Stakeholder Panel on Infant Formula and Adult Nutritionals (SPIFAN) kit samples was performed and compared with parameters defined according to the vitamin D Standard Method Performance Requirements (SMPR (R)). Linearity was demonstrated over the range specified in the SMPR, with the LOD being estimated at below that required. Method spike recovery (vitamin D-2, 97.0-99.2%; and vitamin D-3, 96.0-101.0%) and RSDr (vitamin D-3, 1.5-5.2%) were evaluated and compared favorably with limits in the vitamin D SMPR. Acceptable bias for vitamin D3 was demonstrated against both the certified value for National Institute of Standards and Technology 1849a Standard Reference material (P-(alpha=0.05) = 0.25) and AOAC INTERNATIONAL reference method 2002.05 (p((alpha=0.05)) = 0.09). The method was demonstrated to meet the requirements of the vitamin D SMPR as defined by SPIFAN, and was recently approved for Official First Action status by the AOAC Expert Review Panel on SPIFAN Nutrient Methods.