We undertake a trade-growth accounting exercise by decomposing data on changes in bilateral international trade flows into their direct (endowment accumulation, productivity growth, changes in trade costs, changing preferences) and indirect components (general equilibrium effects). Furthermore, we distinguish between the (potentially partly endogenous) direct and the deep drivers of trade by extending standard new trade models to include endogenous factor accumulation, R&D driven productivity growth, and endogenous changes in expenditures shares on manufacturing goods and services. We quantify the importance of the direct and deep drivers of the growth of bilateral and multilateral trade of 67 economies over 20 consecutive years since 1988 for three trade outcomes: changes in growth rates across different country pairs, changes in levels of trade flows, and changes in trade to GDP ratio. The results suggest that changes in shocks to endogenous endowment accumulation and trade costs have been the most important drivers of trade in the sample at hand, whereas productivity growth and changes in preferences appear to have been quantitatively less important.