Disobedient pupils, off-task behaviour, and bullying are common problems in schools in many countries; they interfere with teaching, create an unsafe learning environment, and challenge the staff. Effective programs involving entire schools to prevent and reduce such problems have already been designed and implemented. However, most interventions target one type of problem behaviour, and their effects have only been evaluated in the short term. The Respect program is broad in the sense that it targets not just one but several types of behaviour in order to prevent and reduce problem behaviour. The program was implemented among all the staff and pupils at three primary schools and one secondary school in Norway. A cohort longitudinal design was used in evaluating the program. Pupils in the four schools reported a decrease in the four areas of problem behaviour. This decrease was sustained or continued after the intervention period for some types of behaviour, even though the results differed between grade levels. In terms of effect size, the results were small to moderate for most grade levels. Although this analysis was non-experimental in nature, it does document sustainable change resulting from intervention in an entire school and suggests that this could be maintained in the long term. (Contains 6 tables, 1 endnote, and 1 appendix.) [The Respect program is partly funded by the Norwegian Directorate of Education and Training.]