If 24-hour esophageal pH monitoring is to be a useful diagnostic tool, it must reliably discriminate gastroesophageal reflux patients despite daily variations in distal esophageal acid exposure. To address this issue, we studied 53 subjects (14 healthy normals, 14 esophagitis patients, and 25 patients with atypical symptoms) with two ambulatory pH tests performed within 10 days of each other. Intrasubject reproducibility of 12 pH parameters to discriminate the presence of abnormal acid reflux was determined. As a group, the parameters of percent time with pH less than 4 (total, upright, recumbent) were most reproducible (80%). Therefore, a subject was defined as having gastroesophageal reflux disease if at least one of these three values were abnormal. Intrasubject reproducibility for the diagnosis of reflux disease was 89% for the entire sample. Among subsets, the reproducibility was 93% for the normals and esophagitis patients and 84% for the atypical symptom patients. Total percent time with pH less than 4 was the single most discriminate pH parameter (85%) and nearly equaled that of the three combined parameters (89%). The intrasubject variability of this parameter was determined by the mean +/- 2 SD of the relative differences between the two test results for all 53 subjects. Total percent time with pH less than 4 may vary between tests by a factor of 3.2-fold or less (218% higher to 69% lower). We conclude: (1) ambulatory 24-hr esophageal monitoring is a reproducible test for the diagnosis of gastroesophageal reflux disease; and (2) the large intrastudy variability in 24-hr total acid exposure may limit this test's usefulness as a measurement of therapeutic improvement.