A process for predicting road noise using virtual prototypes assembled from test-based and CAE-based component models is presented in response to automotive industry trends such as vehicle diversification and electrification, which drive the digitalization of NVH development methods and increased importance of road noise. First, a methodology for test-based component characterization for structureborne road noise is presented using a real-world case study. Then, a virtual prototype is assembled using test-based tire-wheel component models and CAE-based front and rear suspension, subframe and body component models. Finally, the road noise performance of the virtual prototype is predicted and decomposed into contributions from the structureborne paths, which are evaluated from objective and subjective point of view.