Highlights: (1) A 3D simulation model of MWA (microwave ablation) based on the temperature-dependent characteristic parameters and blood flow parameters was established to realize the visual simulation of temperature distribution and coagulation zone. The internal forced convection condition was used to accurately characterize the large vessel. (2) The ex vivo MWA experimental platform was built to verify the accuracy of the simulation model. A peristaltic pump was employed for operatively controlling blood circulation and a medical soft plastic tube was introduced for appropriately simulating a blood vessel. (3) The influences of blood flow parameters of large vessels on temperature distribution and coagulation zone were systematically analyzed in order to provide reference and guidance for MWA clinicians. Purpose: Clinical MWA of liver tumor is significantly limited by the accurate prediction of vascular cooling effects. To provide reference and guidance for clinical MWA of liver tumor, the three-dimensional effects of different blood flow parameters of large vessels on MWA temperature distribution were systematically evaluated. Materials and methods: Firstly, the MWA three-dimensional finite element simulation model with blood flow parameters was established. Secondly, to verify the effectiveness of the model, MWA was performed ex vivo in porcine liver for 360 s and the temperature was measured by thermocouples. A medical soft plastic tube was placed parallel to the MWA antenna to simulate a natural liver vessel. Finally, based on this model, the influences of different vessel diameters and vessel-antenna spacings on MWA temperature distribution were analyzed. Results: Sixteen ablations were performed to verify the accuracy of the simulation model. The mean temperature errors between measured data and simulation results at six measurement points were 3.87 ℃. In the first 10 seconds of MWA, the vessel cooling effect on temperature distribution was negligible. When the vessel-antenna spacing was 5 mm and the vessel diameter varied from 3 mm to 6 mm, the temperature at the measured point near the vessel decreased by 2.11 ℃ at 360 s. When the vessel diameter was 6 mm and the vessel-antenna spacing varied from 5 mm to 7 mm, the temperature at the measured point near the vessel reduced by 14.91 ℃ at 360 s. In addition, blood diameter had little influence on the temperature distribution near the heating point. The volume of coagulation zone will not be obviously affected once the vessel lies outside the predicted coagulation zone. Conclusions: The MWA simulation model with blood flow parameters is established. Vessel-antenna spacing is the primary factor affecting the temperature distribution. A vessel with larger diameter can have a more significant effect on the temperature distribution. The large vessel will take away and block part of conduction heat, so the coagulation zone will not be formed on the lateral side of the vessel.