Summary: A reliability-based optimization method is developed for preform shape design in forging. Forging is a plastic deformation process that transforms a simple shape of workpiece into a predetermined complex shape through a number of intermediate shapes by the application of compressive forces. Traditionally, these intermediate shapes are designed in a deterministic manufacturing domain. In reality, there exist various uncertainties in the forging environment, such as variations in process conditions, billet/die temperatures, and material properties. Randomness in these parameters could lead to variations in product quality and often induce heavy manufacturing losses. In this research, a robust preform design methodology is developed in which various randomnesses in parameters are quantified and incorporated through reliability analysis and uncertainty quantification techniques. The stochastic response surface approach is used to reduce computation time by establishing a relationship between the process performance and shape and random parameters. Finally, reliability-based optimization is utilized for preform shape design of an engine component to improve the product quality and robustness.