We report a new method to measure the fraction of glucose derived from gluconeogenesis using gas chromatography-mass spectrometry and positive chemical ionization. After ingestion of deuterium oxide by subjects, glucose derived from gluconeogenesis is labeled with deuterium. Our calculations of gluconeogenesis are based on measurements of the average enrichment of deuterium on carbon 1, 3, 4, 5, and 6 of glucose and the deuterium enrichment in body water. In a sample from an adult volunteer after ingestion of deuterium oxide, fractional gluconeogenesis using the "average deuterium enrichment method" was 48.3 +/- 0.5% (mean +/- SD) and that with the C-5 hexamethylenetetramine (HMT) method by Landau et al. (Landau BR, Wahren J, Chandramouli V, Schumann WC, Ekberg K, Kalhan SC; J Clin Invest 98: 378-385, 1996) was 46.9 +/- 5.4%. The coefficient of variation of 10 replicate analyses using the new method was 1.0% compared with 11.5% for the C-5 HMT method. In samples derived from an infant receiving total parenteral nutrition, fractional gluconeogenesis was 13.3 +/- 0.3% using the new method and 13.7 +/- 0.8% using the C-5 HMT method. Fractional gluconeogenesis measured in six adult volunteers after 66 h of continuous fasting was 83.7 +/- 2.3% using the new method and 84.2 +/- 5.0% using the C-5 HMT method. In conclusion, the average deuterium enrichment method is simple, highly reproducible, and cost effective. Furthermore, it requires only small blood sample volumes. With the use of an additional tracer, glucose rate of appearance can also be measured during the same analysis. Thus the new method makes measurements of gluconeogenesis available and affordable to large numbers of investigators under conditions of low and high fractional gluconeogenesis ( approximately 10 to approximately 90) in all subject populations.