Abstract: The `Filtering problem' consists of estimating the current state of a dynamical system from a record of noisy measurements. More precisely, consider two stochastic processes {X_n} and {Y_n} with {Y_n} being a random perturbation of {X_n}. We consider {X_n} as an unobservable process and {Y_n} as a measurement process. The `Filtering problem' then consists of computing the conditional expectation of X_n given the observations Y_1,...,Y_n. The asymptotic properties of the filtering process {Z_n} are of great interest both from a theoretical perspective as well as in applications. In this talk, I will report on some preliminary results (joint work with Jochen Brocker) concerning the asymptotic properties of the filtering process when {X_n} is generated by the iterations of an expanding map.