Analysing 137 studies involving over 158,000 participants, this meta-analysis finds that student engagement has large average correlations with both academic achievement (r = .33) and subjective wellbeing (r = .35). Crucially, the relationships differ by engagement type: academic achievement is most strongly linked to behavioural engagement, while wellbeing is most closely associated with affective engagement. The study also highlights a significant conceptual problem — engagement is currently “overgeneralised” as a construct, with inconsistent definitions undermining the quality of research and intervention. An essential methodological reference for understanding what engagement research can and cannot tell us.