The optimization algorithm for redundant 7R serial robot is studied. Use Devavit-Hartenberg matrix to set up a 4 ×4 matrix form closed-form equation of the 7R robot and regard the minimum condition number of the 5 ×5 Jacobian matrix as the control indices for the given 5 constrains of the end-effector in the industry requirement. Change the two joint angles θ 0 and θ 3 and use Newton iteration method to approach and optimize. Seek the minimum condition number of the other five joint angles' Jacobian matrix to complete the redundant optimization and obtain the optimum work status. The algorithm is validated by a group of simulation data.