The Sense of Community Index (SCI) has been the most widely used measure of Psychological Sense of Community (PSC), despite abundant evidence it does not adequately measure its intended dimensions. Alternative model specifications are rarely tested beyond their initial appearance in the literature, and the use of the SCI as a 1‐dimensional scale has been promoted without testing for invariance across groups. We use a sample of students from a small Northeastern state college to test the measurement properties of the SCI and to test for invariance across semesters of college and across waves of panel data. Neither the original 1‐ or 4‐factor solutions to the SCI nor an alternative 4‐factor solution provide an adequate fit to the data. Evidence of measurement invariance was mixed, with invariance found across waves of data, but not across semesters of school. Future research using PSC should employ measures other than the SCI.