According to the USDA’s Foreign Agricultural Service, U.S. farmers export more than 20 percent of what they produce. Since 2000, exports have been rising virtually every year, increasing from $58 billion in 2000 to over $133 billion in 2015. This growth has put pressure on U.S. ports which are vital links to foreign entities and trade. Investment in ports can and do have a significant influence on trade flows. However, there is little research that examines determinants of flows from a port to a foreign country or from the production-rich U.S. interior to the ports en route to foreign countries. In this study, we provide a comprehensive examination of trade for selected agricultural commodities, namely corn, soybeans, wheat, and grain sorghum. This examination includes a description of the countries that import these commodities and the U.S. ports from which they import, as well as a description of the domestic U.S. suppliers of these commodities and the U.S. ports they use to export over time. A focus of the analysis is on the ports that importers and exporters choose to use. Two separate econometric analyses are presented: an “importer analysis” and an “exporter analysis.” In both cases, the analyses focus on the ports used. In the importer analysis (Section I), we model decisions of 151 foreign countries from 96 U.S. ports from 2003 to 2017, and explain which ports are used and the intensity of trade (quantity) between the foreign country and the U.S. port. In the second, exporter analysis (Section 2), we examine the decisions of 70 different origination points in the Upper Midwest and their choice of ports (i.e., where they send their product) from 2014 to 2018. In both cases, the decisions are framed in terms of shipping rates, port identifiers, and port attributes. In both the import demand and export supply analyses, we find that the cost of transportation and port attributes are important variables. The results are then used to evaluate the changes in trade to changes in the transportation costs and in port attributes. In the importer model, the results are restricted to sea-going movements. In the exporter model, ports with a barge option (from movements in the study area) are more likely to be chosen than ports without a barge option, and West coast ports are more likely to be chosen than Northeast ports. The results are also used to measure the responsiveness of decisions to changes in rates and port attributes. Finally, we calculate “willingness to pay” for deeper channels and longer berthing lengths, which are key to evaluating the benefits of investments. In our analysis, the ranking of U.S. ports by total agricultural export tonnage has remained relatively consistent from 2002 to 2017. However, the market share of the top ranked ports has fallen, while the market share of lower ranked ports has grown. The decline in market shares for the top ports and increase in market shares for lower ranked ports indicates that importing countries have diversified the ports used when importing agricultural goods from the United States. More specifically, we find: • Importers of U.S. agricultural products are less likely to choose ports with high associated shipping costs. As freight rates rise, the probability that a port is chosen falls, while the probability that another port is chosen rises. • On the other hand, importers of U.S. agricultural products are more likely to select ports with deeper channels and longer total berthing lengths. • Exporters of U.S. agricultural products are also more likely to ship goods from ports with deeper channels and longer total berthing lengths. Similarly, as freight rates rise, the probability a port is chosen falls. • Competition in barge and rail shipping markets influence how exporters of U.S. agricultural products choose to export. As barge rates rise, the probability of selecting to ship goods by barge to U.S. ports falls and the probability of selecting to ship goods by rail rises.