Researchers have been investigating student dropout, and possible reasons leading to it, for quite a long time. Cumulatively, the results of these research attempts show that drop out of students depends on several factors, of which only a relatively small number can directly be dealt with by universities. Therefore, the overall research problem investigated in this study was to timeously identify risk factors impacting on first-year students' intention to drop out from a university in South Africa in order to manage the potential effect of these factors. By using a cross-sectional quantitative research design, data for the study were collected from first-year students who completed an online survey questionnaire. One of the survey questions requested the respondents to indicate whether they have already (eight weeks after enrolling for the first time) considered dropping out of university; 11.5% of the respondents answered 'Yes' to the question. The data was analysed using partial least squares structural equation modelling (PLS-SEM) to get an indication of the effects (direct and indirect) of constructed latent variables on students' intention to drop out. The outcomes show that only addressing the direct effects leading to drop out (as universities usually do) is not enough and that the indirect effects should also be identified and addressed if one wants to make progress in solving the dropout problem.