Background: Tibial and femoral fractures are among the common fractures and are usually treated with intramedullary nails, being often treated by intramedullary nailing. This study was performed to determine the best preoperative method for estimating the intramedullary nail length.Methods: 82 patients with tibial and 25 patients with femoral shaft fractures were studied for estimating the preoperative intramedullary nail length. The obtained measurement was compared with the length of the suitable nail used at operation. Intraobserver and interobserver reliability of the methods, was assessed by testing on healthy volunteers. For femur fractures the assessed methods were greater trochanter to patella (T-P), greater trochanterlateral epicondyle (T-LE) and olecranon to fifth finger (O-5F) tip. For tibial fractures medial malleolus-tuberosity (MM-T), intact leg radiogram and olecranon fifth metacarpal were assessed. The agreement of the data was assessed by measuring ICC, SEM, CV% and B & A plots.Results: Almost all of the measurement techniqes showed "good" interobserver and intraobserver reliability. In patients with femur fracture, with the greater to patella method showing the best ICC (0.876) and the lowest SEM (0.777). In tibial fracture, the best ICC (0.860) and the lowest SEM (0.602) were for "medial malleolus-tubercle" (MM-T) method. B & A plots were created for the findings of second part of the study, and the conclusion was confirmed. The estimated lengths were larger than the nail lengths used at surgery in 64% of (T-P); 93% in (O-5MF), 100% of (T-LE) for femur; and 40% in (MM-T), 37.3% of (O-MB) and 65.06% of radiograms.Conclusion: The medial malleolus to tuberosity method for tibia and the greater trochanter to patella method for femur fracture showed the best estimated methods; however, we couldn't find a completely reliable method from the all examined methods.