Continuous stirred tank bioreactors with non-monotonic growth kinetics can exhibit bifurcation and multiple equilibrium points. Maximum productivity for this bioreactor is located at the limit of a bifurcation. Therefore, when the desired operating point is near optimal productivity, even small perturbations may drive the system from stability to an undesired, unstable condition. This work describes a fuzzy servo controller that was designed based on sector nonlinearities to force the bioreactor to stay near its optimal productivity point. The designed controller enables tracking of a reference signal that switches between stable and unstable nodes and finally stabilizes at the optimal operating point. Two reference signals were evaluated for the substrate and the biomass respectively. The performance of the proposed controller was compared to PID and LQ controllers. Linear matrix inequalities were used to prove the closed-loop stability. (C) 2012 Elsevier Ltd. All rights reserved.