Objective To investigate the application value of multi-slice CT perfusion in the differentiation of lung lesions. Methods A total of 92 cases lung lesions patients which admitted into the hospital during February 2013 to February 2015 were prospectively analyzed. Among the objects, 58 cases were lung cancer with the diameter of 2-7cm,34 cases were benign lesions, including pulmonary tuberculosis (14 cases), lung abscess (6 cases), inflammatory pseudotumor in (2 cases), nonspecific inflammation(6 cases), hamartoma in (4 cases), and Aspergillus in (2 cases). The diameters of the lesions were 2 to 5cm. All cases underwent multi-slice CT perfusion scanning and the images were transmitted to the workstation and were analyzed by software package. The BV (blood volume), BF (blood flow), PS (surface permeability) and MTT (mean through time) of the lesions were recorded. The application value of multi-slice CT perfusion in the differentiation of lung lesions was analyzed. Results The perfusion parameters like BF, BV, MTT and PS in patients with malignant lesions were significantly higher than those in the benign group, the differences of BV, MTT and PS between the two groups were statistically significant (P<0.05). BV value which was equal to or larger than 6mg/100g was defined as the threshold of malignant lesions. Therefore, the coincidence rate of multi-slice CT perfusion imaging diagnosis in the benign group was 64.71% while in the malignant group was 100.00%. If the PS value which was equal to or more than 30mL/ min·100g was defined as the threshold of malignant lesions, there were 34 30 cases of benign pulmonary lesions which were confirmed as correct,the coincidence rate was 88.24%(30/34). There were 4 cases misdiagnosed (2 cases of pulmonary tuberculosis and 2 cases of lung abscess). In the 58 cases of malignant lesions, there were 50 cases diagnosed, 8 cases misdiagnosed and the coincidence rate was 86.21%. Conclusion The BV and PS values in patients with malignant lung tumors were higher than those in the benign group. CT perfusion imaging can reflect the microvascular characteristics of different types of lung tumors, which is helpful to the differentiation of benign and malignant lung tumors, and is worth of promotion.