Aims: The Drug Education in Victorian Schools (DEVS) programme taught about licit and illicit drugs over two years (2010-2011), with follow up in the third year (2012). It focussed on minimising harm and employed participatory, critical-thinking and skill-focussed pedagogy. This study evaluated the programme's residual effectiveness at follow up in reducing alcohol-related risk and harm. Methods: A cluster-randomised, controlled trial was conducted with a student cohort during years eight (13-14 years old), nine (14-15 years old) and 10 (15-16 years old). Schools were randomly allocated to the DEVS programme (14 schools, n = 1163), or their usual drug education (7 schools, n = 589). Multi-level models were fitted to the data, which were analysed on an intent-to-treat basis. Statistically significant findings: Over the 3 years, there was a greater increase in intervention students' knowledge about drugs, including alcohol. Their alcohol consumption did not increase as much as controls. Their alcohol-related harms decreased, while increasing for controls. There were fewer intervention group risky drinkers, and they reduced their consumption compared to controls. Similarly, harms decreased for intervention group risky drinkers, while increasing for controls. Conclusions: Skill-focussed, harm minimisation drug education can remain effective, subsequent to programme completion, in reducing students' alcohol consumption and harm, even with risky drinkers.