Abstract: Over the past two decades, the use of ideal point estimates---often transformed into measures of preference similarity between countries---has become commonplace in studies of international conflict and cooperation. Similar to the vote scaling approach used to construct measures of ideology for domestic legislators, ideal point estimates are typically constructed from country-votes cast on passage of resolutions on the plenary floor of the United Nations General Assembly (UNGA). In this article, we argue that revealed preferences on UNGA votes are dependent on and vary significantly by the procedural context in which they are cast. We show that constructing such estimates using other votes---those cast in UNGA committees, on amendments, and on procedural matters---yields more precise scores for countries with certain preferences and thus may be more appropriate for answering particular research questions.