Control structure selection plays an irreplaceable role in achieving an effective control system. Through the application of steady-state multiplicities and classic control theories such as zero dynamics and relative gain arrays (RGAs), the primary goal of this article is to select the best control structure from the selected candidates based on heuristic reasoning for a demonstration two-stage-riser catalytic pyrolysis process which can maximize propylene yield without any significant losses in gasoline/diesel yields. Extended bifurcation diagrams and RGAs are adopted to preliminarily evaluate candidates from open-loop perspectives. Closed-loop steady-state multiplicity, which is employed to compare control actions associated with each control structure, can provide a solid basis for the synthesis of control loops. Dynamic simulations for tackling disturbances and tracking set points are finally carried out to evaluate the relevant maximum deviations and settling times under various scenarios, It is illustrated that output temperatures from two risers and the regenerator controlled by two recycled catalyst flow rates and the cooling water flow rate, respectively, which have the best controllability characteristic and most superior dynamic behavior, are the most suitable control structure for further control system design of the studied process. (C) 2014 Elsevier Ltd. All rights reserved.