This study presents a reliability-based design optimization (RBDO) framework for evaluating a four-layered flexible pavement system. The major focus of the study is the optimum design of flexible pavements with respect to fatigue and rutting performance considering the variability associated with design variables. The influence of variability associated with thickness and resilient moduli of flexible pavement layers on fatigue and rutting failures is discussed. Initially, robust nonlinear regression equations were proposed to relate the fatigue and rutting strains to the thickness and resilient moduli of the pavement layers and were validated. Then, the reliability indexes against fatigue and rutting failures for the design of flexible pavements were evaluated using the first-order reliability method (FORM). Typical design charts were presented to estimate reliability indexes against fatigue and rutting based on a wide range of thicknesses and moduli values of pavement layers. The results from the reliability and sensitivity analyses show that the bituminous layer's thickness and resilient modulus of the base layer are the most influential parameters for the fatigue failure. Further, the reliability index against rutting failure is significantly influenced by all the random variables except the bituminous layer's resilient modulus. It is proposed that for pavement systems, a reliability of 95% against both fatigue and rutting failure criteria may be adopted. A reliability-based design example was presented and the results were compared with that of the traditional design method. For the selected mean and standard deviation of resilient moduli of pavement layers, it is noted that the traditional method overestimates the reliability levels by 10%-40% due to nonconsideration of variability associated with the independent layer moduli. © 2019 American Society of Civil Engineers.