This paper studies the linkage between international trade and income convergence across countries. Different theories offer conflicting predictions regarding how they might affect each other. In the existing empirical literature estimating the trade impact on income convergence, a long-lasting problem is the reverse causality from income convergence to trade. This paper provides a disaggregated bilateral trade data analysis to solve this problem. The results show that the reverse causality from income convergence to trade exists in differentiated product sectors, but not in homogeneous product sectors. Trade in homogeneous sectors reduces the income gaps among trade partners, but it is not significantly affected by their income difference. Therefore, the negative effect of trade in homogeneous sectors on the income gap is free from the reverse causality problem. It can be taken as a pure evidence of trade-induced income convergence. This result is robust to various econometric methods.