This study evaluated the school-based short-term, cognitive-behavioral group anger management programme, ‘Learning How to Deal with our Angry Feelings’ (Southampton Psychology Service, 2003). Thirteen groups of children aged 7- to 11-years-old were randomly allocated to two different cohorts: One cohort (n = 35) first received the intervention and was then assigned to a no-intervention period; the other (n = 35) first had a no-intervention period and subsequently received the intervention. Both cohorts showed statistically significant improvements in their understanding of anger directly post-intervention, but neither reported improvements in self-perceived experience of anger scores. Teacher-rated measures of change in some problem behaviors (conduct problems and peer problems) showed significant improvements, but only in the second cohort (no intervention; intervention). The implications of these findings, and possible reasons for differences between cohorts, are discussed.