It is a widely held belief that foreign direct investment (FDI) has a positive effect on economic growth. We test this hypothesis by performing convergence regressions derived from a model of endogenous technological change. We estimate the rate of growth in per-capita income, relative to the per-capita income of the United States, in terms of US FDI, human development, financial development, and trade. We apply a panel approach, instrumenting for explanatory variables and correcting for correlated errors by clustering by countries. The heterogeneity of FDI is taken into account by considering various FDI-related activities – in addition to the conventionally used FDI stocks and flows. Furthermore, we draw on industry-specific FDI data, rather than exclusively on aggregated data. Our empirical analysis puts into question the currently prevailing euphoria about FDI as a means to induce economic catching-up processes of developing countries. We conclude that the central challenge facing policymakers is not to attract FDI, but to improve the local conditions required to benefit from the widely perceived unique advantages of FDI. In addition, our findings support the proposition that FDI stocks do not adequately reflect FDI-related economic activities.