(eng): The paper deals with an output-feedback model predictive control (MPC) for discrete-time systems influenced by bounded disturbances. The proposed MPC combines a state-space design and a state estimation. The state estimates are obtained by a specific uniform Bayesian filter. It provides an evident disturbance attenuation in the estimated state. The MPC design considers a quadratic cost function that incorporates penalties on the tracking error, on the actuation effort and on the system output increments. The theoretical results are completed by illustrative examples using a dynamic model of a parallel kinematic machine as a controlled system.