A method using reduced-order Kalman filters is developed to estimate the thin gas film axial force and moments in real time for mechanical gas face seal systems in a flexibly mounted stator configuration. First-order Gauss–Markov stochastic models, combined with the stator motion equations, form the basis for the reduced-order Kalman filter estimators. Two schemes are presented to estimate axial force and moments based on stator motion measurements. In one scheme, the force and moments are directly estimated and, in another scheme, a set of proper orthogonal decomposition weighting functions is estimated, from which the gas film force and moments are computed. Both estimators are shown to approximate the gas film axial force and moments successfully for different forcing functions over a wide range of compressibility numbers.