In a 6.5 year period starting January 1982, 121 patients (74 male, 47 female; 1.6:1) with complicated gastroesophageal reflux referred to Alberta Children's Hospital, University of Calgary, required a Nissen fundoplication at a mean age of 35.5 months (range 3 weeks to 18 years). The median age of onset of symptoms was less than 1 month. Symptoms and indications for surgery included regurgitation (88%), failure to thrive (52%), reflux-associated pulmonary symptoms and aspiration (48%), biopsy evidence of esophagitis (35%) with heartburn (17%), dysphagia (18%), hematemesis (17%), anemia (13%), and hypoproteinemia (22%). Sixty-four percent of the patients had a syndrome or chromosomal abnormality, respiratory disease, or neuromuscular disorder. The barium contrast upper-gastrointestinal radiographic series, performed in all patients, identified structural [gastric outlet obstruction (2%), esophageal stricture (11%), erosive esophagitis (9%)], and functional abnormalities [gastroesophageal reflux (90%), barium aspiration (8%), esophageal hypoperistalsis (30%), delayed gastric emptying (4%)]. Barium contrast upper gastrointestinal radiographic series identified gastroesophageal reflux with a sensitivity of 90% (compared to history), was 50% sensitive and 92% specific for erosive esophagitis (compared to biopsy), was 59% sensitive and 74% specific for esophageal dysmotility (compared to esophageal manometry), and there was a significant (p less than 0.01) association between barium aspiration and prior evidence of aspiration pneumonitis. Esophageal manometry demonstrated a significantly (p less than 0.001) lower esophageal sphincter pressure in patients compared with controls, but no significant correlation with failure to thrive, aspiration pneumonia, biopsy evidence of esophagitis, or parameters of the 24-hour esophageal pH study. Twenty-four hour pH monitoring showed significantly (p less than 0.05) more reflux episodes than in asymptomatic controls and there was significant (p less than 0.05) correlation between the percentage of time pH was less than 4 and the presence of hypoalbuminemia, and biopsy-proven erosive esophagitis or Barrett's esophagus. Endoscopic appearance was 91% sensitive and 60% specific for esophagitis when compared to biopsy. Nissen fundoplication was completely effective at resolving gastroesophageal reflux in 83%, and associated with marked improvement in 15%. No patient died as a result of fundoplication. Major complications included: recurrence of symptoms requiring reoperation (2%), subsequent mechanical bowel obstruction (8%), wound infection or pneumonia (12%).