In recent years there is a growing interest in determining the impact of inequality on economic growth. Theoretical papers as well as empirical applications have, however, produced controversial results. Although there is a considerable part of the literature that considers inequality detrimental to growth, more recent studies have challenged this result and found a positive effect of inequality on growth. In this paper, we provide a contribution to the empirical puzzle by using meta-analysis to systematically describe, identify and analyse the variation in outcomes of empirical studies. We find that estimation methods, data quality and sample coverage systematically affect the results. The results point out that it will be particularly useful to increasingly focus research on determining the impact of income inequality on economic growth using single-country data at the regional level, or a relatively homogeneous set of countries with adequate controls for country-wide differences in economic, social and institutional characteristics. © Journal compilation © 2008 Scottish Economic Society.