Abstract Excessive static deformations often occur during the end-milling of large thin-walled parts, which play an important role in the aerospace industry. The real-time adjustment of axial cutting depth (ACD) is often adopted to compensate deformation errors. Owing to the coupling effect between the ACD and the force induced deformation, the iterative adjustment of ACD is needed. However, it may not converge with the limited adjustment times especially in high-speed machining with large ACD. We hereby propose an approach to speed up the convergence of the ACD iterative adjustment. Firstly, the real-time full compensation process, which is commonly used in the existing real-time compensation methods, is mathematically analyzed and proved to be only first-order convergent. Then, an approach based on the secant method, which provides a higher order of convergence, is developed and introduced into the real-time compensation process for large thin-walled plate machining. Finally, the superiority and effectiveness of the proposed approach are demonstrated through an ACD iterative adjustment simulation and an experiment of cutting square pockets.