A linear optimal control problem with an H-infinity constraint can be transformed into a finite optimization problem, with the number of variables bounded in terms of the problem’s data.