The innervation patterns of the rectal pouch and fistula of 52 children with anorectal malformations were investigated. Posterior sagittal anorectoplasty was used for intermediate and high anomalies; for the latter it was combined with an abdominal approach. Perineoproctoplasty was performed for low anomalies. The specimens were investigated by acetylcholinesterase staining, lactate dehydrogenase, and succinyldehydrogenase reaction. They consisted of fistula material only in 23 patients and of parts of the rectal pouch in 29. Fourty-four patients (84.6%) had follow-up, and information of bowel movements and continence was obtained after a mean of 3.3 years. Abnormal innervation patterns were found in 96% of the specimens. All fistulas were found to be aganglionic, including the adjacent part of the rectum involving the internal sphincter equivalent. Classical aganglionosis was found in 31% of the rectal pouch specimens, hypoganglionosis in 38%, neuronal intestinal dysplasia (NID) type B in 14%, and dysganglionosis in 10%. All patients with severe constipation or soiling at the time of follow-up had some histopathological correlation. Of the 25 patients for whom the specimens had consisted of rectal pouch material, nine (31%) had severe constipation. All four patients with a low-type malformation who had follow-up and pathological innervation patterns in the rectal pouch suffered from severe constipation; this was true of only five of the 19 children with intermediate or high malformations (P < .05). However, numerous pathological innervation patterns had been identified in patients who had normal bowel function at the time of follow-up. It is concluded that partial denervation of the rectum may not be the only cause in the pathogenesis of constipation after posterior sagittal anorectoplasty and perineoproctoplasty. The high frequency of neuronal intestinal malformations in the rectal pouch may be related to the higher frequency of bowel disturbances in patients with low malformations, in whom the resection was less radical. However, the clinical course is not necessarily related to specific histopathological findings. In the authors' opinion, the recommendation to use the distal rectal pouch and parts of the fistula in the reconstruction of anorectal malformations should be reconsidered.