Lithium-ion battery has advantages of high energy density and cost effectiveness than other types of batteries. However due to the low mechanical stability, their performance is strongly influenced by environmental conditions. Especially, external pressure on a cell surface is a crucial factor because an appropriate force can improve battery cycle life, but excessive force may cause structural failure. In addition, battery pack is composed of various components so that uncertainties in dimension and material properties of each component can cause a wide variance in initial pressure. Therefore, it is important to optimize structural design of battery pack to ensure initial pressure in an effective range. In this paper, target stiffness of module structure was determined based on cell level cycle life test, then structural design has been optimized for weight reduction. Cell cycling tests were performed under different stiffness conditions and analyzed with regression model. Based on that model, target stiffness of module structure was set to 12,000kgf/mm to achieve cycle life requirements. For weight reduction of module structure, various design optimization techniques were conducted and compared. Deterministic design optimization (DO), reliability-based design optimization (RBDO) and progressive meta-model (PMM) based optimization methods were used and the reliability analysis was performed with Monte-Carlo simulation, First- and Second- order reliability method (FORM and SORM). Weight and reliability (SORM) of initial design were have been improved from 1.062kg and 48.6% to 0.746kg and 50.2% with DO, and 1.015kg and 99.9% with RBDO. With PMM method, the weight can be reduced to 0.926kg with the lower computation cost. Validation test was performed on pack level with optimized module structure and 2/3C cycling. Optimized battery pack showed capacity fade of 12.7% at EOL which is 39.6% of improvement compare to initial designed pack. By adopting proposed reliability analysis and design technique, the cycle life and weight compatibility of the battery system can be improved.