A model-free iterative learning control strategy (ILC) in nonrepetitive trajectories, applied to robotic manipulators is presented in this article. The development of this ILC controller is derived in general case for second-order nonlinear multi-input multi-output (MIMO) systems subjected to external disturbances. Unlike most of the ILC implementations in the literature, which considers both the resetting condition and periodic trajectories; the proposed ILC controller is model-free and works under the practical alignment condition and noncyclic desired trajectories. Furthermore, these trajectories may vary in magnitude and in the starting and terminal points. The Lyapunov stability approach is combined with a time scale transformation to synthesize the control law. The asymptotic convergence of both the position and velocity tracking errors are demonstrated along the iteration axis. Finally, experimental results in pick and place operations of a parallel Delta robot are presented and discussed. These results are very effective and clearly point out the high potential of the developed model-free ILC technique for industrial robotic applications.