The logic of the ‘smile curve’ in the context of global value chains (GVCs) has been widely used in case studies of individual firms, but rarely identified at the country-industry level by using real data. This paper puts forward a proposal, based on an inter-country input–output model, to consistently measure both the gain of value added and the position of countries and industries when they join GVCs. This allows for better identification and mapping of economy-wide smile curves in a given conceptual value chain. Using the World Input- Output Tables, we identify the Information and Communications Technology exports- related smile curves for China and the United States (US), which provide an intuitive and visual representation of who gains value added and jobs through joining GVCs, and to what extent. Further insight into the distributional implications of GVC expansion, based on our analysis of labour markets for China and the US, provides a strong support for the so-called ‘Paradoxical Pair of Concerns’ between developed and developing countries. Our empirical results show that gains through joining GVCs may vary greatly across different skill levels of labour domestically, a fact that has, at least in part, been a driver of the backlash against globalization and the rise of trade protectionism.