The present study implements a time-optimal control strategy for a discontinuous aerobic bioreactor, used to treat highly concentrated toxic wastewater present in some effluents of the chemical and petrochemical industries, using respirometric techniques. The control strategy regulates the feed rate to maintain a constant optimal substrate concentration in the reactor, which in turn minimizes the reaction time. Since this control requires on-line knowledge of unmeasurable variables, an Extended Kalman Filter is used as a nonlinear observer. The experimental setup was a 7 litre laboratory bioreactor used to treat synthetic wastewater with high concentrations of 4-chlorophenol. The controller consisted of a personal computer with data acquisition hardware and real-time software tools, peristaltic pumps and an electronic oxygen meter. Three experiments were performed: one to obtain parameters and calibrate the observer, another one to validate the time-optimal strategy and a final one to evaluate theperformance of a fully automated time-optimal operation. When well calibrated, the observer provided good enough estimates and the controller worked as expected, reducing reaction time and increasing the overall efficiency of the bioreactor, when compared with the usual SBR-type operation.