To evaluate leading mechanisms of fat embolism, we investigated patients who received femoral procedures prospectively. Forty patients were classified into three groups according to reaming procedures. Group A (10 cases) was the control group who operated with open reduction and internal fixation for peritrochanteric fractures. Group B (18 cases) was composed of hip arthroplasty patients and group C (12 cases) patients received intramedullary nailing procedures due to femoral shaft fracture. Pulmonary function test and chest X-ray were taken to evaluate pulmonary condition. Perfusion and inhalation radionuclide scan proved pulmonary embolism and fat embolism. Arterial blood gas analysis, lipid metabolites (cholesterol, triglyceride and NEFA (Non-esterified Fatty Acid)) and coagulation battery (PT, aPTT, TAT-III (Thrombin anti-thrombin III complex) and D-dimer) were measured in preoperative, intraoperative (reaming and post-reaming) and postoperative period, individually. Although two cases of fat embolism were verified only in group C preoperatively, there was no difference with the incidence of pulmonary embolism between reaming and non-reaming group (P<0.05). Blood gas analysis revealed no difference between control group and reaming one in each stage. NEFA in lipid metabolism was abruptly increased in group C only during reaming procedure. The titer of D-dimer and TAT-III were elevated specifically in fat embolism patients by reaming procedure. We suggest that D-dimer and TAT-III may be preliminary screening tools for the early diagnosis of fat embolism. And the coagulation system seems to be more related to fat embolism rather than lipid metabolism.