The natural environment is associated with better behaviour and academic performance in children. However, research to date has been cross-sectional and it is important for experimental studies to investigate if a causal relationship exists. Further, participant samples from areas characterised as disadvantaged are underrepresented in the research. This study investigates the effect that lessons in nature have on disadvantaged young people’s behaviour and learning compared to lessons in a standard classroom over one school term (10 weeks). A quasi-experimental study was conducted in a socioeconomically disadvantaged city in Australia. Three classes were taught in a standard indoor classroom for 5 weeks; then two of those classes relocated to a green outdoor classroom for the remaining 5 weeks; researcher observations of redirect rates, teacher ratings of behaviour, and academic grades were analysed. Students (13–14 years old) spent more time on-task in outdoor classrooms at the rate of an extra 20 s per teacher redirect, and this effect lasted over several weeks. Although engagement was better, this did not translate to better grades; reasons for this discrepancy are discussed. Outdoor classrooms may promote less disengagement and misbehaviour in class, and this could be significant for those affected disproportionately by it.