An optimal control design method is introduced and then applied to the optimum design of active and passive suspension systems. A basic three-dimensional 7-DOF car riding model is considered. The method allows the arbitrary choice of sensors for various state variables to be used for feedback control of each suspension unit. Fully-active, limited-active, passive, and two versions of semi-active systems are studied and compared.The results indicate that the dominant feedback gains that have the most impact on the suspension performance measures are those which are proportional to locally measured signals. Also, when no tire deflection measurements are available there may be a complete coincidence between the behaviour of active systems that use different sets of measured variables.