We retrospectively reviewed a prospectively followed cohort of 105 patients with 110 open tibia fractures treated with external fixator or intramedullary nail to determine whether smoking affects fracture healing. Severe open tibia-shaft fractures treated at a tertiary-care medical center were included. Patients with type II, IIIA, or IIIB tibia fractures were eligible. Treatment for all patients was similar, except that they were randomized to receive external fixator or intramedullary nail. Time to fracture healing was the main outcome measurement. Smokers had a union rate of 84% (52/62), and nonsmokers had a union rate of 94% (45/48), P = .10. For smokers in one arm of the study, time to union was significantly longer (P = .01), and there were more complications (P = .04). Smoking decreased unions, slowed healing, and increased complications.