The problem of model matching by state feedback is reconsidered and some of the latest results are discussed.