Generally, industry equipment is composed of several subsystems, and criticality of the subsystems in the equipment are different so that assigning reliability in design phase, confirming the detection level in running phase, selecting the maintenance policy for each subsystem must be based on its criticality. The improved TOPSIS used to evaluating the criticality of each subsystem of industry equipment satisfies the requirement of practical engineering. TOPSIS is employed to evaluate the criticality after the weights of each attribute are assigned using the information-entropy method. It is shown by the instance——criticality analysis of plastic extruder that the proposed method is effective and applied for evaluating the equipment criticality.