Résumé : Background. The 13C urea breath test (13C-UBT) is the most convenient method for diagnosing Helicobacter pylori infection noninvasively. Nondispersive isotope selective infrared spectrometry (NDIRS) is an inexpensive and easy alternative to mass spectrometry. The objective of this study was to evaluate: (1) the reproducibility of the 13C-UBT as performed by using the NDIRS method; (2) the repeatability of bags analysis and the impact of delayed analysis; and (3) the need for fasting status for the 13C-UBT. Methods. The 13C-UBT was performed with 75 mg urea labeled with 13C, with breath samples collected at times 0 and 30 minutes. Results are expressed as delta over baseline (0/00). Fifty-three patients underwent two successive 13C-UBTs with an interval of 48 to 72 hours. The 106 collected bags were randomly reanalyzed immediately or 72 hours later. In 26 volunteer subjects, the 13C-UBT was performed both in a fasting condition and after a nonstandardized meal. The reproducibility was assessed by the method of Bland and Altman. Results. The mean of difference between two successive tests was 0.14 0/00 (standard deviation, 0.90), and the coefficient of repeatability was 1.80 (confidence interval, 95%). The difference between two successive analyses was always less than 2.2% of the initial value. The coefficient of variation between two successive tests for the influence of a meal was 11.24. Conclusion. The 13C-UBT as performed by using NDIRS is reproducible, analyses can be delayed up to 72 hours, and the test must be performed in fasting conditions.