International business cycles are synchronized across developed economies. To understand this empirical phenomenon, I develop a multi-country real business cycle model with trade that embeds a comprehensive set of shocks. I match the data exactly with the endogenous outcomes of the model so that shocks fully account for the data, including GDP and bilateral trade shares. Quantifying the model to a panel of G7 countries and the rest of the world during 1992–2014, I find that shocks to bilateral trade linkages are largely responsible for the business cycle synchronization. In contrast, country-specific correlated shocks play relatively minor roles. Additionally, I find that trade-linkage shocks help to resolve the trade co-movement puzzle, by predicting a much stronger link between trade and cross-country GDP correlations.