Seventy-one extended multiple-level esophageal pH studies were performed in 61 infants and children suspected of having gastroesophageal reflux (GER). The patients were placed in one of six clinical groups based on their presenting symptomatology. They were classified as "normals" or "refluxers" based on a reflux score calculated from a pediatric modification of the Johnson and DeMeester distal esophageal pH criteria. For the total group, excluding postoperative patients, all differences noted in the distal esophagus between normals and refluxers persisted at the middle and proximal esophageal levels (P less than .001). Significant differences between normal and reflux patients were noted at all levels in the central nervous system damage and postoperative groups. Only the middle esophageal probe detected differences for all variables in the failure to thrive group (P less than .01). In the esophageal atresia group the middle level sensor detected differences in all parameters studied and the proximal sensor noted differences in three of the four parameters (P less than .01). Significant differences between normals and refluxers in the apnea group were best demonstrated in the proximal esophagus. There was no benefit in monitoring the pH of the more proximal esophageal levels in patients with esophagitis. Extended four-channel esophageal pH studies were highly effective in detecting clinically important GER. The patterns of reflux in infants and children vary markedly with the presenting symptoms. Distal esophageal pH monitoring may not demonstrate evidence of GER that would otherwise be detected by the middle and proximal esophageal sensors. This new ambulatory outpatient technique of pH monitoring in young patients suspected of having GER provides advantages over the conventional esophageal pH study.