Duodenogastric reflux (DGR) is a natural event, occurring occasionally and whose pathological significance is not well known. The accuracy of 24-hour gastric pH-metry for the DGR diagnosis has been tested in adults but not in children. For this purpose we measured the area under curve (AUC) at pH 4, 6 and 7 and the percentage of total time above pH 4 (%pH4), 6 (%pH6) and 7 (%pH7), excluding the 2-hours post-prandial period in 88 children suspected of having gastroesophageal reflux (GER). Forty were considered normal whereas 40 had acid GER and 8 with GER. In the control group %pH4 was 9.68 +/- 14.1, %pH6 3.4 +/- 8.7 and %pH7 1.0 +/- 2.4. The values for AUC were 101 +/- 112.8, 16.5 +/- 22.3 and 3.3 +/- 6.1 pH unit/min, respectively. Comparable results were found in the acid GER group. On the other hand alkaline refluxes had higher figures for all parameters: %pH4 22.1 +/- 13.9, %pH6 12 +/- 13 y %pH7 7 +/- 12, AUC at pH 4,406.5 +/- 410, 136.1 +/- 194 at pH6 and 48.2 +/- 85.1 at pH7 (p < 0.05). Because of the large dispersion of values in the control group we selected the 95 percentile, as the upper limit of normal values instead of the mean +/- SD. Therefore the upper limit were 27 for %pH4, 9 for %pH6 and 3.8 for %pH7. The AUC, 316, 64 and 16 pH/min, respectively. These results prove that DGR in children is a very common event and confirm that DGR definitely contributes to alkaline GER.