We performed a prospective, randomised trial in 39 patients with open tibial fractures treated initially by external fixation to compare cast immobilisation (group A) and intramedullary nailing (group B) as a sequential protocol planned from the onset of treatment. The results showed that group B achieved faster union (p <
0.05) than group A with less malunion or shortening and a greater range of movement. Patients treated by intramedullary nailing required fewer radiographs and outpatient visits (p = 0.0015) and had a more predictable and rapid return to full function. We feel that these severe fractures are better treated by delayed intramedullary nailing and that this has an acceptable rate of complications.