Despite the significant evidence base demonstrating the positive impact of the Promoting Alternative Thinking Strategies (PATHS) curriculum on children's social-emotional and mental health outcomes, there has been very little research on its efficacy in improving academic attainment. More generally, the relationship between implementation variability and PATHS intervention outcomes has been underexplored. A cluster-randomised controlled trial with two arms: intervention (PATHS -- 23 schools) and control (usual practice -- 22 schools) was implemented to assess the impact of PATHS on English and Mathematics for children in years 5 (aged 9-10 years, n = 1705 pupils) and 6 (aged 10-11 years, n = 1631 pupils) in English primary schools. Two-level hierarchical linear models (school, child) were used to assess both primary 'intention-to-treat' effects and secondary 'subgroup' effects (for children eligible for free school meals). Additionally, the moderating role of implementation variability was assessed in 31 year 5 (n = 712 pupils) and 32 year 6 (n = 732 pupils) classes across the 23 intervention schools, with fidelity, dosage, quality/responsiveness and reach data generated via classroom-level structured lesson observations. Intention-to-treat and subgroup analyses revealed no significant positive effect of PATHS on children's academic attainment. Cluster analyses of observational data revealed four distinct implementation profiles, differentiated primarily by dosage levels. However, these profiles were not significantly associated with differential academic outcomes. In light of our findings and their likely generalisability, it is not possible to recommend PATHS as an effective intervention for improving the academic attainment of children in English primary schools.