Objectives: Mindfulness-based programmes (MBPs) show potential for universal prevention in schools, given that they target shared risk factors underpinning common mental health disorders. High-quality research in youth remains limited, but suggests that MBPs may be less effective in early- versus mid-adolescents. Methods: This randomised controlled trial tested the acceptability of an 8-week MBP in year 8 (Mage 13.7) and year 10 (Mage 15.5) students (N = 434; 48.2% female) and compared outcomes (mindfulness, depression, anxiety, eating disorder risk factors, wellbeing) to normal curricular controls. Results: Levels of acceptability were moderate and did not differ by age band. For younger students, there were no differences at post-intervention for mindfulness students compared to controls (Cohen’s d <.22) but at 3-month follow-up, the mindfulness group were worse in wellbeing (Cohen’s d = −.25; 95% CI −.49, −.01) and two aspects of mindfulness (Awareness of External Environment, d = −.30; −.55, −.06; Decentering and Nonreactivity, d = −.39; −.63, −.15). For older students, there were no significant differences between the intervention and control groups at post-intervention (Cohen’s d <.17) or follow-up (d <.22). Conclusions: We suggest that the use of formal meditation in unscreened conscript early teens may be unwise. Further research is needed to identify acceptable and effective age appropriate modifications of MBPs for early- and mid-adolescents before they can be supported as universal interventions in schools. Suggestions are made for ongoing research in this area.