Abstract: The acoustic release is an essential device in the field of marine engineering, and its reliability is of paramount importance. Traditional reliability analysis primarily relies on static predictions, which fail to accurately reflect the real-world impact of performance degradation under prolonged extreme conditions. To address this, this study adopts a state-estimation-based reliability research method for deep-sea acoustic release devices. First, the entire lifecycle of the release device is systematically analyzed, and its mission profile is thoroughly examined to establish a reliability testing model. Second, the study conducts theoretical analysis and reliability prediction of the acoustic release device's mission profile. Given the characteristics of long-term deep-sea值守, the battery pack is identified as a key degrading component affecting system reliability. Therefore, with battery health state (SOH) estimation at the core, a simulation-based SOH estimation model and capacity decay curve are developed to classify reliability levels and predict remaining useful life (RUL). Simulation results demonstrate that state estimation of critical components such as the battery enables the transition from static prediction to dynamic assessment of reliability. The proposed method not only provides a theoretical foundation for lifespan prediction of deep-sea acoustic release devices but can also be applied to reliability studies of other critical marine equipment.