In this paper we show that inconclusive results in previous empirical studies on human capital and growth might be due to omitted variable bias. Using data for about 130 countries, we show that after inclusion of variables related to the social capabilities concept of Abramovitz (1986) i.e. economic opportunities and quality of legal institutions, the human capital variable turns out to be significant. We also show that economic opportunities significantly moderate the relationship between human capital and growth. The results are robust to different variants of indices for economic opportunities and the quality of legal system.