Regurgitation and aspiration of feedings is a significant problem in children with impaired oral intake fed via gastrostomy. Using extended (18-24 hour) esophageal pH monitoring to assess gastroesophageal reflux (GER), we studied prospectively 32 children (aged 2 to 16 years) referred for feeding gastrostomy. Twenty-five patients had repeat esophageal pH monitoring after surgery. Prior to surgery, GER was documented in 23 (72%) of the 32 children. Twenty-two of the 23 children with GER before surgery had an antireflux operation performed in conjunction with the feeding gastrostomy. Gastroesophageal reflux was clinically significant in the single failed antireflux operation and in the child with GER before surgery who only had a gastrostomy performed. All nine patients without GER only had gastrostomy performed. Six of these developed GER by pH monitoring after surgery, with significant vomiting in four. Of our 11 patients remaining at risk for GER after surgery, seven (64%) had persistent vomiting with gastrostomy feedings. Thus, 91% (29 of 32) of the children were potentially at risk for GER if a gastrostomy only was performed. We believe these data support the need for a "protective" antireflux operation in children referred for feeding gastrostomy.