A value iteration approach based solely on input/output measurements is proposed to solve linear quadratic (LQ) optimal control problems for single-input, single-output (SISO) continuous-time systems.