Abstract： Objective To investigate the value of time-intensity curve of dynamic contrast enhancement MR imaging in the discrimination of benign and malignancy in musculoskeletal tumors. Methods Ninety patients were examined with fast acquisition with muhiphase enhanced fast GRE series. The TIC of lesions were obtained using slope images in which pixel intensity reflected the slope value. The curves were classified according to their shapes as type Ⅰ , washout enhancement; type Ⅱ, plateau enhancement; type Ⅲ, gradual enhancement. Taking pathological diagnosis as gold standard, the power of the maximal enhancement slope and curve types in discriminating benign and malignant lesions was evaluated by appropriate statistic analysis. Results There were 49 malignant and 44 benign lesions. The distribution of curve types for malignant tumors was type Ⅰ 75.5% ( 37/49), type Ⅱ 24. 5% (12/49). While the numbers for benign tumors was type Ⅰ 59. 1% ( 26/44 ), type Ⅱ 15.9% ( 7/44 ) and type Ⅲ 25.0% ( 11/44 ), respectively. The patterns of curve types in malignant lesions were different from benign lesions significantly ( χ2 = 14. 008, P < 0. 01 ). The slope value in benign lesion was 6. 80 + 3. 35 and that in malignant lesion was 6. 80±2. 71. The difference was not statistically significant( t = 0. 008, P > 0. 05 ). Type Ⅰ and type Ⅱ (excluding lesions with typical benign morphology ) were suggestive of malignant tumors. Type Ⅲ was indicator of a benign lesion. The diagnostic indices for the shape of TIC criterion were: sensitivity 100%, specificity 50%, positive predictive value 78%, negative predictive value 100% and accuracy 82%, respectively. Conclusion Combined with the characteristic of morphology, the TIC improves the power of MR imaging in discriminating benign from malignant musculoskeletal tumors.