Much of the research on economic stratification at selective colleges relies on data with limitations that tend to restrict how comprehensively or accurately studies can assess the incomes of students enrolled at selective universities, particularly over time. Studies that use quality data tend to find that the share of students at selective colleges who are low income has remained remarkably stable since the turn of the century. In this report, the authors set out to address some of the limitations in the literature on enrollment at selective universities and test the popular narratives related to this topic. The authors use a data set that few researchers have enlisted for this type of analysis, the National Postsecondary Student Aid Study, and define selective colleges as the 200 most selective public and private institutions nationally. The authors also conduct a separate analysis for public flagship universities. They do not find evidence that the share of students enrolled at these 200 institutions who are from the lowest income quartile declined during the years covered in the study. Students from high-income families were a growing share of enrollment at these institutions in the mid-2000s. Meanwhile, the share of students at selective colleges who are from middle-income families has steadily declined over time, particularly students from the third income quartile.