Student enrollment and retention are high priorities for universities these days, and Stephen F. Austin State University (SFA) is no exception. Schools are increasingly using geographic analysis methods to improve enrollment and retention. One particular method being used to track and predict enrollment is the gravity model (a.k.a. spatial interaction model). This study has the first known gravity models of retention rates or Texas enrollment. This study presents two gravity models, one each for freshman enrollment and freshman retention rates at SFA during 2001-2005. The models use as predictors public school graduates, competition from other colleges and universities, and distance from SFA. The models were linearized using logarithms and calibrated using ordinary least squares regression. The enrollment model was highly significant, and the residuals were analyzed. The retention model was not significant. In its place, a scatterplot of freshman enrollment vs. retention rates was analyzed at three different enrollment levels.