Economic planning, when considered as a control problem, has as its objective establishing optimal (or sub-optimal) trajectories for the variables subject to Government Control. This means that the policy variables (control), instead of being arbitrarily determined by the policymakers, will be the result of an optimization process, with the objective of reaching pre-established goals. In this work a Certainly Equivalence Adaptative Control is applied to a macroeconometric model of the Brazilian economy with measurement error. Since the employment of time-invariant models has been widely criticized (Lucas critique) the model used here is time- varying. An adequate way to treat such a case is through an adaptative control scheme, in which control and identification of the model are perfomed simultaneously. By means of simulations the policy obtained with the adaptative controller is compared to the policy adopted by the Brazilian Government.