Under thermal abuse conditions, lithium-ion batteries are subject to multiple sources of uncertainty, which can potentially trigger thermal runaway. To enable reliable structural design under thermal safety constraints, this study proposes a reliability-based design optimization (RBDO) method for lithium-ion batteries based on critical venting prediction. First, an analytical model is developed that couples electrochemical reactions, heat conduction, gas dynamics, and nonlinear elasticity, enabling a comprehensive characterization of the thermogas-mechanical evolution, with the critical venting time adopted as the performance metric. Second, global sensitivity analysis using a variance-based decomposition method identifies high-sensitivity parameters for dimensionality reduction. Finally, an RBDO model with venting response probability as the constraint is formulated, and an efficient solution strategy is established by integrating the performance measure approach with a decoupled optimization frame-work to ensure computational efficiency and numerical stability. Experimental results show that the proposed model achieves prediction errors lower than 3% for temperature and critical response. Compared to existing methods, it achieves a superior balance between accuracy and efficiency, with RBDO solution time under two hours. The proposed approach demonstrates strong engineering applicability and extensibility, offering an effective tool for safety-oriented structural optimization of lithium-ion batteries.