A solution to the robust optimal control problem, subject to local states constraints in the feedback law, is presented. Examples with matched and unmatched uncertainty in the system matrix are presented. Results obtained through simulation are shown as a means of validating the proposal.