A local analysis of stability and observability of the ethanol process are determined for continuous operation. It is found that the process is stable and noncompletely observable in the selected steady states. The analysis of observability was studied in terms of the observability matrix rank conditions. Furthermore, we design a nonlinear observer in order to estimate the observable state variables. The software sensor (state estimator) is developed considering only one measurable variable, the glucose concentration, and taking in account model uncertainties, which is a realistic issue. Some sufficient conditions for the existence of the proposed observer are obtained, which guarantee the convergence of the proposed methodology. The maximum ethanol production conditions are obtained by manipulating the dilution rate with optimal initial substrate concentration and the observable subspace is determinate which allows estimating six state variables (starch concentration, susceptible starch concentration, ethanol concentration, biomass concentration, glucose concentration and enzyme concentration). Numerical simulations are provided to show the effectiveness of the proposed observer where a comparison with a standard sliding-mode observer is done. (C) 2012 Elsevier Ltd. All rights reserved.