A gas chromatography-mass spectrometry method for determining exposure to the chemical warfare agent sulfur mustard (SM) has been developed. The technique is based upon quantifying thiodiglycol (TDG) released from blood protein adducts that are formed upon exposure to SM. Protein was precipitated from plasma, whole blood, or packed red blood cells and then treated with sodium hydroxide to liberate protein-bound TDG. The TDG was derivatized with pentafluorobenzoyl chloride that enabled sensitive detection by negative-ion chemical ionization. Octadeuterothiodiglycol was used as an internal standard to account for compound loss due to the sample work-up procedure. Exposure of human plasma to SM (12 nM to 200 nM) resulted in a linear relationship (r2 = 0.9958) between SM concentration and released TDG levels with means ranging from 5 to 76 pg/mg protein. The coefficients of variation expressed as a percentage for the data points ranged from 2 to 38%. The application of this procedure was demonstrated in two SM animal exposure models. African green monkeys (Chlorocebus aethiops) were exposed intravenously (iv) to 1 mg/kg SM and TDG levels in blood samples were analyzed out to 45 days post-exposure. Mean TDG levels were determined to be 220 pg/mg protein on day one and declined to 10 pg/mg protein on day 45. Yorkshire cross pigs (Sus scrofa) were cutaneously exposed to neat liquid SM and TDG levels in plasma were determined out to 21 days following exposure. Mean TDG levels were found to be 60 pg/mg protein on day one and decreased to an average of 4 pg/mg protein on day 21. The data from this study indicate that the assay is sensitive, reproducible, precise, and provides a relatively simple approach to assay TDG cleaved from blood proteins at relatively long time frames (21- 45 days) after SM exposure. The utility of the method has been demonstrated in vivo in non-human primate and pig SM exposure models.