In this paper we present two approaches for designing H-2-suboptimal stable controllers. Both full-order and reduced-order controllers are considered.