The aim of this study is to explore the existence of different segments of future university students on the basis of their university selection criteria and to analyse the extent to which the segments identified differ as regards the students' perceptions of the reliability of university communication tools. This has been done using 605 usable questionnaires, while the segmentation method employed was a latent class cluster analysis. Future students were divided into five clusters based on their university selection criteria: (1) high academic performance; (2) high academic performance but economy and word of mouth (WOM); (3) unconcerned; (4) independence from parents; and (5) overinformed. The differences in the perceptions of the reliability of universities' communication tools among the clusters identified were studied using the bias-adjusted three-step approach. Students in the cluster labelled 'high academic performance but economy and WOM' tended to rate the university's communication tools as more reliable, while those included in the cluster labelled 'independence from parents' found them to be less reliable. This paper contributes to the marketing theory for higher education by modelling the heterogeneity in the students' choices. This makes it possible to cover one of the main gaps in the literature on higher education.