Due to the high pace of development in the automotive industry, there is a need for innovating cost engineering. A methodology for intelligent cost estimation in the early stages of the product life cycle is introduced. In a first step it is shown how significant economic and technical parameters for cost prediction can be prepared and filtered from historical calculation data. Subsequently, it is shown how cost prediction models can be developed using machine learning algorithms. Learning data and practical use cases come from a large automotive manufacturer in Germany. The models predict the costs of car parts and assemblies of increasing complexity. Seven different machine learning models are trained and optimized. Based on the test data of the use cases these models are assessed and compared. Finally, the prediction results obtained are evaluated from different perspectives, demonstrating the practical applicability of the most suitable methods explored.