The distortion incurred along a shipyard panel line is a costly factor of shipbuilding influenced by design parameters and manufacturing operations. By identifying the most impactful parameters, design for production can effectively mitigate panel line distortion. Bayesian networks can model the potential for distortion as a result of specific design parameters and process operations. Bayesian networks capture a complex web of cause-effect relationships, such as the mechanics of a structural panel’s production, and break them down into a series of smaller conditional relationships that are more easily established. By breaking down a panel line and creating a series of connected networks modeling each step, the potential for distortion can be predicted from design parameters such as plate thickness and cutout locations and manufacturing decisions such as the use of automated seam welding. The series of networks allows the distortion measured upstream to update the probabilities of downstream nodes. Determining the effect of panel parameters on the distortion generated along a panel line allows informed detail design decisions that could reduce costly distortion in the shipyard. This paper explores the capability of Bayesian networks to predict such distortion through demonstration on an active shipyard panel line.