Abstract: Abstract: A self-cleaning screen filter has broad application on the domestic and foreign markets. The variation law of head loss, and the relationship between filtering time and self-cleaning time have been analyzed through different experiments. However, the flow distribution of filter inside cannot be obtained by traditional experiment methods. To fully understand the internal flow structure and characteristics of a self-cleaning screen filter and find the inadequacies of flow field distribution, the whole flow was simulated through Fluent (6.3). Based on the analysis of filter structure and self-cleaning principle, the mathematical model of inside flow field and the dynamic grid model of a self-cleaning system were established respectively, the computing area and the grid of filtering and self-cleaning processes, as well as boundary condition of inlet and outlet were proposed. In order to ensure the reliability of the simulation results, the relationship between the flow rate and pressure of a self-cleaning system were compared in this study, and the results show that the simulated pressure agreed with the measured values. The internal flow field of the filtering process was simulated, and the distribution law of flow velocity, turbulent energy, and average pressure were studied. The flow velocity decreases along the radial direction of the filter body, and it decreases rapidly from the inlet value of 2.6~3.0 m/s to the outlet value of 1.4~1.8m/s. The maximum average pressure was located at the inlet of flow, and the average pressure of the first filtering chamber was significantly higher than the value of the second filtering chamber, which the average pressure reduces relatively 25% and 80% after the flow passes the first and second filtering chambers. As the distributions of flow and turbulent energy are extremely uneven, the clogging of the entire screen is uneven. The design of inlet and outlet should be changed in order to avoid the uneven clogging of the entire screen, for example, several small-diameter screens replace a big diameter-screen. Moreover, through the technology of dynamic mesh, the process of self-cleaning was also simulated, and the flow fields of flow velocity, turbulent energy and average pressure were analyzed. The inlet of sunction will have a huge attractive power, which can ensure that the sediment of the screen surface be cleaned. But the velocity, average pressure, and turbulent energy of sunction vary greatly at different heights of sunction. The maximum and minimum values of velocity are located respectively at upper and lower sunction, and the maximum velocity difference is about 8.5m/s, which results in the lower screen being cleaned unthoroughly and the upper screen being damaged easily. The pressure gradient near the sunction is large, and the maximum pressure difference between outside and inside the sunction is about 1.1MPa. The maximum and minimum values of average pressure are located respectively at middle and upper sunction, which results in different cleaning efficiency at different heights of sunction. The maximum average value of turbulent energy is located at lower sunction, and a local high turbulent energy area also appeared at upper sunction. The simulated results of a self-cleaning system indicate that the distribution of sunction tubes can not be uniform along the axial direction, and the position of suction tubes can be adjusted and the number can be increased, which can increase the sunction power of lower sunction tubes and improve the cleaning efficiency of sediment.