This article presents a filtered-dynamic-inversion controller for multi-input multi-output linear minimum-phase systems. The controller requires limited model information, specifically, knowledge of the relative degree and the first nonzero Markov parameter. Furthermore, the controller is effective for stabilisation, command following and rejection of unknown-and-unmeasured disturbances. This article analyses the stability and closed-loop performance (i.e. command following error) of filtered dynamic inversion, which is a single-parameter high-parameter-stabilising controller. We show that for sufficiently large parameter the closed-loop is asymptotically stable, and the average power of the performance signal is arbitrarily small.