In this communication, a nonlinear observer based control system is developed and illustrated by a reactive batch distillation column (RBDC) producing ethyl acetate. This ethyl acetate is the product of the esterification reaction between acetic acid and ethanol. Being the lightest component, ethyl acetate comes out first as distillate from the top of the column. With the aim to achieve a constant product purity, a nonlinear control scheme is developed, consisting of an extended generic model controller (EGMC) and an observer for the estimation of states required for controller simulation. Here two nonlinear state observers, namely neuro extended Kalman filter (NEKF) and neuro high gain observer (NHGO) are devised. Both the NHGO and NEKF are basically coupled with an artificial neural network (ANN) model, which aims to infer the product composition based on the measured tray temperature required for the closed-loop high gain observer (HGO) and nonlinear extended Kalman filter (EKF), respectively. Through open-loop simulation study, performance of these two state observers is evaluated. Although both the state observers are good enough to track the process output, the NEKF proves its superiority over the NHGO. In the subsequent closed-loop constant composition control investigation, a comparison is made between the EGMC-NHGO and EGMC-NEKF with respect to a traditional proportional integral (PI) controller. In this closed-loop comparative study, it is observed that the performance of the EGMC-NEKF is better than the EGMC-NHGO and the conventional PI. (C) 2018 Elsevier Ltd. All rights reserved.