In this study, we use decomposition methods on PISA 2006 data to compare student academic performance across OECD countries. We first establish an empirical model to explain the variation in academic performance across individuals, and then use the Oaxaca-Blinder decomposition method to decompose the achievement gap between each of the OECD countries and the OECD average. Results indicate that the explained portion of the achievement gap varies across countries. In some countries, our empirical models are able to account for almost all the achievement gap, while unexplained country-specific effects still dominate in other countries. Finally, we use two Asian countries, Japan and Korea, to demonstrate how to identify major factors that have contributed to the observed achievement gap across countries.