Although fewer adolescents are consuming alcohol than was the case in previous decades, those who are consuming alcohol are still exposed to alcohol-related harms. While the evidence for the effectiveness of universal, school-based interventions is limited, a recent cluster randomised controlled trial (The STAMPP Trial) reported a significant effect at 10 months post-intervention of a combined classroom/parental intervention on heavy episodic drinking (HED) in the previous 30 days, but no significant effect on the number of self-reported alcohol-related harms (ARH) experienced in the previous 6 months. This follow-up study sought to examine intervention effects 24 months after delivery of the intervention (+ 57 months from baseline, or + 34 months post-intervention). Participants were 5029 high school students in STAMPP (38% of 12,738 pupils originally randomised into the trial), from 87 schools (82.3% of schools recruited in the original STAMPP trial). Outcomes were assessed using two-level random intercepts models (logistic regression for HED and negative binomial for number of ARH). Results of the present study show that the intervention effect for HED deteriorated over the following 2 years (OR declined from 0.60 to 0.97), and there was still no difference in ARH. This was due to an increase in the prevalence of intervention students' HED rather than a reduction in prevalence in control students. Results are discussed in the context of prevention initiatives.