Relative merits of three methods of diaphragmatic hernia repair were evaluated in growing animals. Twenty-five puppies underwent laparotomy. In four controls, the left hemidiaphragm was incised and sutured primarily. In the remaining dogs, it was partially resected sparing the phrenic nerve. The defects were repaired in six with silastic sheeting, in eight with polytetrafluoroethylene (PTFE; trademark, Gore-Tex), and in seven with a thoracoabdominal muscle flap. Dogs were killed at 1, 4, and 7 months for gross and microscopic evaluation of the repair. Diaphragmatic function was evaluated by inspiratory force against a closed airway and by selective phrenic nerve stimulation (PNS). Serial fluoroscopy was used to evaluate diaphragmatic motion. Grossly the diaphragms in all groups showed compensatory growth. Microscopically the silastic was encapsulated without adherence, while PTFE showed tissue ingrowth. Maximal inspiratory force was equivalent in all groups but selective PNS revealed left-sided impairment in all experimental groups. Fluoroscopy showed paradoxical motion of the diaphragm in the muscle flap group for 1 to 2 months, and in the silastic repair group for 2 to 3 weeks, with near normal motion in the PTFE group for the entire postoperative period. These differences disappeared by 6 months. Prosthetic materials or muscle flaps are all safe for repair of large diaphragmatic hernias. Diaphragmatic growth occurs and the prosthesis remains in place. Physiologic impairment is minimal and not of clinical importance. Use of PTFE may be the preferred method as it develops better tissue incorporation and results in more normal diaphragmatic motion in the critical early postoperative period.