The rotary water jetting is one of the most important techniques for horizontal well cleanup. The jet flow is used to remove plugging particles from sand control screens to recover their permeability. Currently, the operation optimization of this technique depends mainly on experience due to absence of applicable evaluation and design models for removing plugging materials. This paper presents an experimental setup to simulate the cleanup process of plugged screens by rotary water jetting on the surface and to evaluate the performance of a jetting tool. Using real plugged screens pulled from damaged wells, a series of tests were performed, and the qualitative relationships between the cleanup efficiency and various operational parameters, such as the type of fluids used, flow rate, mode of tool movement, etc., were obtained. The test results indicated that the cleanup performance was much better when the rotary jetting tool moved and stopped periodically for a certain time than that when it reciprocated at a constant speed. To be exact, it was desirable for the rotary jetting tool to move for 1.5-2 m and stop for 2-4 min, which was called the “move-stop-move” mode. Good cleanup performance could be obtained at high flow rates, and the flow rate was recommended to be no lower than 550-600 L/min. The test results also indicated that complex mud acid was better than clean water in terms of cleanup performance. Good cleanup efficiency and high screen permeability recovery could be achieved for severely plugged screens. Rotary jetting is preferred for the cleanup of horizontal wells with severely plugged screens, and the screen permeability recovery ratio may reach 20% if optimized operation parameters were used.