This study investigates the fatigue behavior of electroactive polymer (EAP) membrane actuators under coupled electromechanical loading to enhance the reliability and durability of EAP-based devices in smart applications. The motivation of this study is to address fatigue failure in EAP membranes, which are increasingly used in soft robotics, artificial muscles, and adaptive structures, yet frequently experience premature failure under dynamic loading conditions. The research employs a viscoelasticity neo-Hookean model to simulate the mechanical behavior of EAP membrane actuators. Based on the principles of crack nucleation and configurational mechanics, the three principal configurational stresses of the model are calculated. A fatigue life factor is introduced to evaluate the fatigue state of the membrane at different positions. The investigation focuses on two key factors influencing fatigue behavior: the elastic polymer network ratio and pre-stretch level. The study systematically analyzes the fatigue increment of the membrane over time under both constant and half-sine cyclic loading conditions. Simulations demonstrate that appropriate pre-stretching significantly improves fatigue resistance of EAP membrane actuators under both constant and half-sine cyclic loading conditions, identifying an optimal pre-stretch level. Additionally, the study reveals that the elastic polymer network ratio plays a crucial role in determining fatigue behavior, with higher network ratios generally leading to improved fatigue performance. These findings inform the design and application of EAP-based devices. By providing insights into fatigue mechanisms and offering strategies to mitigate fatigue failure, this study contributes to the development of more reliable and durable soft actuators. The results can be applied to optimize EAP membrane performance in various smart systems including soft robotics, wearable devices, and adaptive structures.