Substantial evidence suggests that countries with stronger trade linkages have more synchro- nized business cycles. The standard international business cycle framework cannot replicate this finding, uncovering the trade-comovement puzzle. In this paper we investigate the extent to which more sophisticated trade models can sort out this puzzle. We show that under certain macro-level conditions but irrespective of the micro-level assumptions concerning trade (within a large class of trade models) synchronization is explained by three factors: (i) the correlation between each country's productivity shocks, (ii) the correlation between each country's share of expenditure on domestic goods, and (iii) the correlation between each country's productivity and the partner's share of expenditure on domestic goods. An empirical investigation of the link between trade and each of the three factors shows that the trade-comovement relation is explained by the first and second factors.