This paper represents the development of a new design methodology based on data mining theory for decision making in vehicle crashworthy components (or parts) development. The new methodology allows exploring the big crash simulation dataset to discover the underlying complicated relationships between vehicle crash responses and design variables at multi-levels, and deriving design rules based on the whole vehicle safety requirements to make decisions towards the component and sub-component level design. The method to be developed will resolve the issue of existing design approaches for vehicle crashworthiness, i.e. limited information exploring capability from big datasets, which may hamper the decision making and lead to a nonoptimal design. A preliminary design case study is presented to demonstrate the performance of the new method. This method will have direct impacts on improving vehicle safety design and can readily be applied to other complex systems.