In this paper, we study a convexified robust control problem and its relation to gain-scheduling control. It reveals that the robust control synthesis condition becomes convex under a special plant structure. Moreover, for this class of robust control problems, the gain-scheduling control approach from scaled small-gain theorem will not provide any performance improvement over convexified robust controllers. Based on this observation, a convexified robust control synthesis framework with stringent performance and computational efficacy is proposed. (C) 2004 Elsevier Ltd. All rights reserved.