<p>The automotive industry, especially in the competitive electric mobility market, faces increasing pressure to quickly develop new car models while maintaining or raising quality standards. Therefore, it is imperative to ensure the efficient production of high-quality car bodies, as they determine the downstream compatibility in production as well as the vehicle’s geometric appearance to end consumers and crucial safety functions. In common practice, an experience-based process is employed to reach the desired geometry of assemblies in car body production. This process necessitates extensive iterative loops that are both costly and time-consuming. This work proposes a method for data-based geometry predictions of car body assemblies. Four steps are defined, which include the analysis of the initial state of the production, the data acquisition process and the development of the model. Its applicability is verified through experimental validation using components from high-volume series production. The results show that by utilizing only geometric production data it is possible to forecast the geometry of car body assemblies.</p>