The purpose of this study is to verify the clinical efficacy of unreamed nails compared to reamed nails. The cases of 31 parients in whom 32 fractures of the femoral shaft had been treated by intramedullary nailing with reamed or unreamed nails were prospectively reviewed. Reamed nailing was done for 17 cases, and unreamed nailing was done for 15 cases. The following results were obtained. 1. According to Winquist-Hansen classification, type 1 was 9 cases, type 2,3,4 were 4,2,2 cases 2. The average operation time was 70.5 minutes in reamed group, and 62.5 minutes in unreamed group(p-value=0.638). the average bleeding amount was 450ml in reamed group, and 218ml in unreamed group(p-value=0.000). 3. The mean times to partial weight bearing and full weight bearing were 4.4 weeks, 7.7 weeks in reamed group retrospectively, and 7.2 weeks, 13.7 weeks in unreamed group(p-walue=0.039,0.002). 4. The mean union time was 15.2 weeks in reamed group, and 17.7 weeks in unreamed group(p-value=0.237). 5. According to Denker's functional classification, 15 cases were excellent, 1 case was satisfactory, 1 case was very poor in reamed group, and 13 cases were wxcellent, 1 case was satisfactory, 1 case was poor in unreamed group(p-value=0.545).