International trade data are filled with discrepancies–where two countries report different values of trade with each other. I develop a novel trade data quality index for reconciling the discrepancies in bilateral trade data. I calculate the quality for each country’s imports and exports separately for every year from 1962 to 2016 and reconcile international trade data by picking the value reported by the country with higher data quality in every bilateral flow. The reconciled data reshape our views on international trade: (a) countries with low data quality under-report their imports and exports: low-quality reporters are 14% more open to trade using reconciled data; (b) corruption, the level of development, and erroneous reporting can explain data quality; (c) importers’ data are more accurate; (d) China tends to under-report its exports and over-report its imports, while there is only a small difference between US self-reported and reconciled data.