This paper presents a gravity model of student migration flows to higher education institutions (HEIs) in Ireland. The analysis is performed on a novel dataset containing detailed information on a range of push' and pull' factors, allowing one to estimate the effects of a number of important school-level characteristics on these flows. This is achieved by estimating and comparing a fixed-effects Poisson model and two conditional fixed-effects negative binomial models and selecting the best model on the basis of the Akaike information criterion (AIC). The preferred negative binomial model accounts for over-dispersion in the student flow data and allows for estimation of the parameter coefficients of the HEI-invariant characteristics. The analysis suggests that while geography plays a very important role in explaining student flows, so too do a range of school-level characteristics. Furthermore, it is found that distance has a differential impact across HEIs and HEI types with important implications for policy-makers and HEI managers.