All H-2 optimal measurement-feedback controllers are characterized and parameterized and then systematic methods of designing H-2 optimal controllers with estimator-based architecture are presented. A method of assigning the closed-loop poles at more desirable locations while still preserving H-2 optimality is also given.