The Catofin propane process is one of the leading industrial processes for production of propylene from propane. It is composed of multiple adiabatic fixed-bed reactors which undergo cyclic operations where dehydrogenation of propane and regeneration of catalyst alternate over roughly ten minutes of period for each. One of the major concerns in the operation of the Catofin process is to maintain the reactor at an optimum condition overcoming gradual catalyst deactivation. Aiming at this, online optimization combined with repetitive control has been investigated. The optimizer computes optimum initial bed temperatures for the dehydrogenation period. The repetitive controller performs cycle-wise feedback action during the regeneration period to attain the target bed temperatures. Both the optimizer and controller were designed on the basis of a first-principle reactor model updated in real-time. Numerical study has shown that the proposed online optimizing control system performs satisfactorily coping with catalyst deactivation and other disturbances.