Open femur fractures in children are uncommon and usually associated with other injuries. In adults, there is a current trend to treat open fractures with intramedullary (IM) devices. The goal of this study was to compare external fixator (EF) to IM devices in the treatment of open femur fractures in children. Diaphyseal femur fractures without growth plate involvement were included. Thirty-five patients (12 IM; 23 EF) were identified. Age, hospital stay, polytrauma, mechanism of injury, and Gustilo-Anderson grade were recorded. Follow-up was at least until the fracture was clinically and radiographically healed. Patients with EFs were 5.2 times more likely (95% confidence interval, 1.05-25.5) to have any complication. Excluding pin track infections, patients with EFs were 2.7 times as likely (95% confidence interval, 0.567-13.2) to have a complication. Refractures occurred only in the EF group (6/23, 26%) and not in the IM nailing group (P = 0.062, Fischer exact test). These were associated with varus malunions-all 3 of the EF group with more than 15 degrees of varus at fracture union suffered a refracture. Treatment of open femur fractures in children is a challenging problem. Treatment with IM devices had fewer complications than the EF. We think that whenever possible, the use of IM devices for the treatment of open femur fracture in children should be considered, especially grade 1 open injuries. If EFs are used, avoiding varus malunion may decrease the refracture rate, and secondary change to an IM device should be considered. Comparative cohort study. Grade 3 level of evidence.