A dynamical output feedback control for the regulation of the chemical oxygen demand (COD) in anaerobic digestion is discussed in this paper. The control law can have different structures: (i) nonlinear if a nominal value of the influent COD concentration is available, and (ii) linear if such a nominal value is not available. The control law includes a high-gain observer used as the dynamic estimator, which allows one to achieve the COD regulation despite modeling errors (which are mainly related to the kinetic terms). However, the dynamic estimator can induce undesired effects such as the so-called peaking phenomena. Such a phenomenon produces large overshoots on the control input and leads to windup behaviour as a result of constraints on the control input. To improve the performance, two schemes are proposed: (a) an antiwindup scheme to consider the constraints in the control input and (b) a fuzzy-based gain-scheduling scheme to tune online the control parameters. Both schemes are designed for the nonlinear and linear structures of the controller. The performances of the proposed schemes are compared by means of a performance index via numerical simulations.