We reconsider the question of what determines corruption at the cross-national level, using new methods and data: observations of occurrences of cross-national corruption. We find that economic development and a small population is associated with lower levels of corruption, as are freedom of the press, political rights, the presence of established democratic institutions, the salience of women's role in society, and low exports of natural resources such as oil. The particular structure of the data also allows for the first time to consider the "relational aspects" of corrupt relationships, which come to the fore when parties to the corrupt transaction, the briber and the bribee, reside in different countries. Overall, we find limited evidence that the relational factors that we consider affect corruption, beyond the effects that they often have on bilateral trade.