In this study, the state-derivative feedback LQR design is proposed for the active vibration control of a vehicle suspension system. In active vibration control problems, state derivative signals which are velocity and acceleration signals are more accurately obtained rather than displacement and velocity signals, since the accelerometers are mostly used. Therefore, state derivative feedback control has been employed instead of state feedback control which is frequently applied in the active suspension literature. Controller design is expressed as a convex optimization problem with linear matrix inequalities. A five-degree-of-freedom model including suspension, seat and driver dynamics is used to examine controller performance in terms of ride comfort, safety and power consumption. Proposed controller and state feedback LQR structure have been examined for different road roughness grades and driving velocities by the use of ISO2631 standard.