We studied 309 constipated infants and children who underwent rectal manometry (n = 261) and/or punch rectal biopsy, in which acetylcholinesterase histochemistry (n = 124) was compared with routine histologic examination (n = 93) for diagnostic accuracy. A 99% diagnostic accuracy was achieved with acetylcholinesterase histochemistry compared with 61% accuracy with routine hematoxylin-eosin staining on punch rectal biopsy. We recommend screening cooperative, quiet patients with rectal manometry. Most patients with abnormal manometric studies or those on whom manometry cannot be performed should have superficial punch biopsy (except those ill infants who require urgent operative intervention). This plan of management provides the least invasive but the most accurate evaluation of the constipated child. Deep rectal biopsy can be avoided except in the very rare child with hypoganglionosis or "skip-area" aganglionosis, which can be detected by manometry.