This paper presents a cross-coupling (CC) controller for tracking contours to effectively reduce the contouring error of a direct-drive robot. Because contouring performance is a primary target over point-to-point tracking performance in a contour-tracking task, a cross-coupling control (CCC) algorithm design based on Lyapunov stability criteria and the recursive updating technique is proposed to enhance the contouring performance by coordinating the motion of multiple axes in spite of considerable model uncertainties and external disturbances. Furthermore, the proposed CCC design, which is a typical MIMO system with nonlinear time varying characteristic, has been verified as being locally stable. The control algorithm developed in tracking a circular and an elliptic contour is experimentally implemented on a two-axis direct-drive manipulator for various payload configurations. It is seen that the controller exhibits a certain degree of robustness, with the contouring performance being only slightly affected by changes in the payload. A comparison of the experimental results with those obtained by a tracking control law, which has no contouring consideration, and a conventional PID-type CCC tuned with the learning automata technique indicates that the proposed CCC can significantly enhance the contouring performance under different contouring commands and various payload configurations.