Aims: Concepts that may be effective in alleviating negative alcohol-related consequences among college students are protective behavioural strategies (PBSs), such as alternating alcoholic and non-alcoholic drinks. Our study used an innovative hypothesis-testing statistical technique (structural equation modelling, SEM) to examine the relationship of PBSs, drinking behaviour and school performance among college student drinkers. Methods: We used confirmatory factor analysis and SEM to explore the relationship between drinking behaviour, PBSs and school performance among a sample of college students (N = 371) who reported drinking alcohol in the past 30 days. Findings: Results indicated that PBSs were positively associated with school performance and drinking behaviour mediated the relationship. Based on the combinatory rules for standardized root mean square residual (SRMR) and comparative fit index (CFI), this model represented a good fit to the data ([chi][superscript 2] = 34.4, df = 17, SRMR = 0.03 and CFI = 0.98). Conclusions: Our study extends the line of research examining the relationship between drinking behaviour, including PBSs, and academic performance of college students by showing the usefulness of an innovative hypothesis-testing statistical technique--SEM. Researchers might want to consider utilizing SEM when examining such relationships in the future. In addition, based on the results of our analyses, colleges and universities with an interest in the academic performance of their students may want to consider implementing and promoting programmes that concern increasing the awareness of PBSs. (Contains 1 table and 3 figures.)