This is a mixed method study carried out from a pragmatist philosophical position. The research question is how (if at all) do primary school pupils change following a residential adventure education experience, how does any change relate to their experience during the residential and what implications does that have for the provision of residential adventure education? It is a three phase study. Phase 1 is quantitative: a survey to assess whether there is a correlation between the extent of residential opportunities and whole school performance measures. Phase 2 is qualitative: a series of interviews with headteachers, parents and policy makers to discover their perceptions of the impact of a residential programme. Phase 3 is quantitative: designing and testing an instrument to measure the impact on pupils of different aspects of a residential programme and comparing this with their classroom attainment and their social and emotional development. No relationship was found between the extent of residential opportunities and whole school performance measures, although it was found that opportunities are inversely correlated with deprivation. Interview data produced a rich source of evidence for the way in which different aspects of a course combine together to produce a powerful impact. Complexity theory was used as a theoretical perspective to suggest that a non-linear step change in self-confidence could arise naturally and possibly inevitably as a result of the fact that residential adventure education is a complex system. Analysis of the pupil impact survey showed that many different aspects of the experience combine to create the impact but that it can reliably be separated into four components: living with others, challenge, teacher relationships and learning about self. There was a significant correlation between the improvement in individual pupils’ classroom attainment over the course of a term and the impact that the residential had on them. There was a significant improvement from pre-course to post-course in pupils’ prosocial behaviour and a significant reduction in perceived hyperactivity.