In the above paper (Fang and Chow, 1998 [Iterative learning control of linear discrete-time systems multivariable system. Automatica, 34, 1459-1462]), a new iterative learning control rule is proposed for linear discrete-time multivariable systems. In this correspondence a counterexample to the Fang and Chow's control rule is presented. It is shown that the results given in their paper are incorrect.