In the automotive industry, high competition, volatile markets and short development cycles are commonplace. In addition, the transition to new drive technologies brings further challenges and complexity for car manufacturers. Improved product cost estimation especially during the early development phases enhances design decisions to better address this situation. Due to the lack of data availability in this stage, exact calculations are very difficult and require a lot of experience. Even though various methods to estimate the product cost already exist, they usually vary greatly in accuracy and applicability. Therefore, the focus of this paper is to investigate different qualitative and quantitative product cost estimation methods in the early development phases. First, a literature review provides the fundamental understanding about existing cost estimation methods. Aiming for a data triangulation, documents were collected, experts interviewed and a survey conducted, all at one engineering company dealing with powertrain development. Based on these three sources, two empirical studies were performed (study A: documents and interviews; study B: survey). A subsequent comparison allowed statements about the applicability of qualitative and quantitative cost estimation methods in the early development phases. Finally, two of these methods (case based and fuzzy logic) were applied at two different points in the early stage of battery pack development. In order to represent the increasing data availability when progressing in the development process, different parameters were included. While the reference data for the product cost estimation came from literature, the case company provided the input data for six battery packs of existing electric vehicles. This allowed a comparison of the estimated cost to the actual cost, whereas the calculation of derivations built the basis to evaluate the accuracy of the methods at both points.