Abstract: To address the issue of the identification of position-independent geometric errors (PIGEs) of rotary axes in dual-swiveling head five-axis machine tools being susceptible to repeatability errors, a novel PIGEs identification method based on the pose data of the machine tool's end-effector is proposed. A mapping model between the PIGEs of the rotary axes and the end-effector pose error is established, and the least squares method is utilized to identify each error parameter. The influence of the machine tool's repeatability error on the accuracy of parameter identification is also analyzed. Simulation results demonstrate that, compared to the existing identification method based on the two-step measurement of the tool center point, the proposed method using end-effector pose data is less affected by the machine tool's repeatability error. The identification accuracy of specific error items is significantly improved, with a maximum of 95%. And the detection efficiency is also higher.