The purpose of this study was to evaluate soft tissue sarcomas by dynamic (18)F-FDG-PET studies, and to establish an index of kinetic parameters for evaluation of their malignancy, histological grade and prognosis, after surgical resection. One hundred and seventeen patients including 79 with histologically proven soft tissue malignancies, 14 with primary benign soft tissue tumors and 24 with postoperative scar tissues were examined. The (18)F-FDG studies were accomplished as a dynamic series for 60 min. The evaluation of the (18)F-FDG kinetics was performed using the following parameters: standardized uptake value (SUV), global influx (Ki), computation of transport constants (k1-k4) with consideration of the vascular fraction (VB) according to a two tissue compartment model, and fractal dimension (FD) based on the box-counting procedure (non-compartmental model). Discriminant analysis (DA) was used for data evaluation. Multivariate analysis was performed to assess the predictive value of each kinetic parameter on survival. Our results showed that in the primary cases (n=46), SUV, k1, Ki and FD were higher in sarcomas than benign tumors. The diagnostic sensitivity of 62.50%, a specificity of 92.86%, and an accuracy of 71.74% were achieved by using the combination of k1 and SUV as input variables for DA. In the postoperative cases (n=71), SUV, VB, k3, Ki, and FD were higher in recurrent lesions than in scar tissues. DA revealed a sensitivity of 80.85%, a specificity of 87.50%, and an accuracy of 83.10% by using the combination of SUV, Ki and FD. In liposarcoma patients (n=32), SUV and FD were higher in GII,III tumors as compared with GI. DA led to a sensitivity of 86.96%, a specificity of 55.56%, and an accuracy of 78.13% by using the combination of SUV and FD. By multivariate analysis of primary soft tissue sarcomas (n=26) after surgical resection, groups with k3>0.025 (P<0.0026) or FD>1.25 (P<0.0162) had significantly poor prognosis. In conclusion, the evaluation of full (18)F-FDG kinetics provides important information for the diagnosis of malignant lesions, histological grading and prognosis of soft tissue sarcomas.