We propose a theory-based approach to testing the presence of the 'home market effect' in a multi-country world. Our framework extends Krugman's (1980, Am. Econ. Rev. 70(5), 950-959) model, in which the appeal of a country as a production site depends on both the relative size of its domestic market ('attraction') and its relative proximity to foreign markets ('accessibility'). We show that the extended model predicts a home market effect only after the actual production and trade data have been corrected for the impact of countries' differential access to world markets. This can be achieved through a simple theory-based linear filter. We propose a series of non-parametric sign- and rank-tests that are closely related to those used in factor proportions theory. When applied to a cross-section of OECD and non-OECD countries, the filtered data performs better than the raw one and our results strongly support the presence of home market effects.