In this paper, we develope a direct adaptive control framework for linear discrete-time uncertain MIMO systems, based on assumptions that the system is controllable and the difference of the unknown system matrix from the stable solution is bounded by a given value. The proposed framework is Lyapunov-based, and the controller guarantees adaptive stabilization. In addition, the adaptive laws are characterized by means of Kronecker calculus. Furthermore, the results can easily be extended to time-varying cases, where the deviation from the stable solution is bounded, and the controllability assumption holds. The controller not only tolerates plant deviation, but also guarantees asymptotical stability of the closed-loop system. Two numerical examples are provided to demonstrate the performance of the controller.