In this manuscript the state estimator design for a class of continuous bioreactor with Liouvillian system characteristics is addressed. The design procedure take into account the sampled nature of the measured output for this kind of processes, considering the model systems structure, the state equations are transformed in a sampled (discrete) system version, which under the considered assumptions are possible to estimate state variables employing only current and past measured output values and model information, without tuning parameters to yield an exact and immediate state estimation. The corresponding theory results of the proposed estimator are applied, where numerical experiments illustrate a satisfactory performance in comparison with a non-linear Luenberger observer. (c) 2006 Elsevier B.V.. All rights reserved.