This paper aims to examine the implications of using the GTAP database and three databases containing information from trade in value-added (TiVA) datasets for computable general equilibrium (CGE) experiments. A comparison of experiments based on the GTAP data and on USITC research on global value chains (GVC) illustrates that a CGE model specified in such a way as to better reflect the trade linkages found in modern global supply and value chains can produce substantial differences in macro level impacts and also reflect the realities of specific product chain relationships. Our comparison would expand to include TiVA datasets from the WIOD project and OECD/WTO.