We present preliminary results on subsonic cavity flow control using reduced-order model based feedback control derived from experimental measurements. The reduced-order model was developed using the Proper Orthogonal Decomposition of PIV results in conjunction with the Galerkin projection of the Navier-Stokes equations onto the resulting spatial eigenfunctions. The stochastic estimation method was used for real-time estimate of the model time coefficients from dynamic surface pressure measurements. Equilibrium analysis led to the linearization of the reduced-order model around the equilibrium point and a model for controller design was obtained by shifting the origin of the coordinates to the equilibrium point. A linear-quadratic optimal controller was then designed and tested in the experiments. The results obtained are very promising and show that control is capable of reducing the cavity flow resonance not only at the Mach 0.3 flow, for which the reducedorder model was specifically derived, but also at other flows with some variation of the Mach number. These preliminary results indicate that the control switches the flow from a single mode resonance to a multi-mode resonance.