International trade in food is an essential component of the global food system, with consequences ranging from environmental sustainability to public health. Evidence-based food policy requires analysis and interpretation of trade flows among countries. We describe a stepwise mixed-methods process to identify and correct errors in international trade data to develop a reliable food trade database for the Pacific region. The method profoundly changes estimates of regional food trade. Similar results would likely be identified in other global regions with trade data quality challenges. If so, improved data quality could have significant food and other policy ramifications.