In order to assess the results of the AO unreamed femoral nail (URFN), and specifically its effects on healing, 147 consecutive patients treated were reviewed. These included 50 reamed femoral nails (RFN) and 97 unreamed femoral nails. Exclusion of pathological fractures, revisions and fractures outside the femoral diaphysis left 51 procedures in which the healing process could be studied. Twenty-four unreamed and 27 reamed femoral nails in patients with diaphyseal fractures AO (32) were followed up by clinical review and radiographically until union or death. There were two deaths from multiple injuries (one in each group) and two non-unions (at 52 weeks), one in each group. There were no cases of infection, angular deformity of leg length discrepancy; two cases required early rotational correction. There was a single broken distal locking screw in the URFN group but no other implant failures. The fractures in the URFN cases took longer to heal with a mean of 26.9 weeks as opposed to 20.5 weeks in the RFN group (P = 0.009). This did not cause a significant clinical problem. The URFN proved easy to use with a much shorter operation time.