A method for determining the site and extent of deuterium (D) labeling of glucose by GC/MS and mass fragmentography was developed. Under chemical and electron impact ionization, ion clusters m/z 328, 242, 217, 212, and 187 of glucose aldonitrile pentaacetate and m/z 331 and 169 of pentaacetate derivative were produced. From the mass spectra of 13C- and D-labeled reference compounds, glucose carbon and hydrogen (C-H) positions included in these fragments were deduced to be m/z 328 = C1-C6, 2,3,4,5,6,6-H6; m/z 331 = C1-C6, 1,2,3,4,5,6,6-H7; m/z 169 = C1-C6, 1,3,4,5,6,6-H6; m/z 187 = C3-C6, 3,4,5,6,6-H5; m/z 212 = C1-C5, 2,3,4,5-H4; m/z 217 = C4-C6, 4,5,6,6-H4; and m/z 242 = C1-C4, 2,3,4-H3. After correction for isotope discrimination and deuterium-hydrogen exchange, the D enrichment of these fragments can be quantitated using selective ion monitoring, and the D enrichment of all C-H positions can be obtained by the difference in enrichment of the corresponding ion pairs. The validity of this approach was tested by examining D enrichment of known mixtures of 1-d1-, 2-d1-, 3-d1-, and 5,6,6-d3-glucose with unlabeled glucose and D enrichment of perdeuterated glucose using these fragments. This method was used to determine deuterium incorporation in C1 through C6 of blood glucose in fasted (24 h) rats infused with deuterated water. The distribution of deuterium was similar to that found by Postle and Bloxham (1980, Biochem. J. 192, 65-73). Approximately one deuterium atom was incorporated into C5 and only 75% deuterium atom was incorporated into C2. The enrichment of C2 and C6 of glucose relative to that of water indicated that 74 +/- 9% of plasma glucose was newly formed 4 h after the onset of deuterium infusion, and gluconeogenesis accounted for about 76 +/- 7% of the glucose 6-phosphate flux.