Background: Interventions to reduce health inequalities can either be targeted interventions (focused on disadvantaged groups and designed to close a health inequalities gap) or universal interventions (available to all and designed to reduce a health inequalities gradient). Other interventions which do not explicitly aim to reduce health inequalities may also reduce, maintain, or increase gaps or gradients in health. Few systematic reviews have explicitly sought to measure the impact of preventive interventions on health inequalities where all three categories of intervention could be included. Objectives: To conduct a systematic review of the impact of secondary school-based cognitive behavioural type (CBT) interventions to prevent poor mental health in young people. To test the feasibility of applying an ‘equity lens’ in a meta-analysis in order to examine whether gaps and gradients in mental health were reduced, maintained or increased. Methods: We conducted a systematic review of RCTs, employing standardised and tested methods for meta-analysis, sub-group analysis and meta-regression to analyse data. A coding scheme captured pre-defined socio-demographic categories implicated in the broader determinants of health. Results: Meta-analysis of 17 high quality RCTs showed a reduction in symptoms of depression and anxiety, which was generally short term. Studies reported no useable data with which to evaluate the differential impact of interventions according to differences in participants’ gender, age, religion, education or social capital. Although conclusions about impact of CBT on inequalities are tentative, data suggested that CBT may be less effective for people who are more socio-economically disadvantaged. Conclusions: CBT delivered to young people in secondary schools can reduce the symptoms of depression and anxiety. Better and fuller reporting of socio-demographic data of participants in reports of primary research is required. Future trials of health promotion interventions should consider including sub-group analyses of population groups at risk of inequalities on specified health measures.