It has long been assumed that schools which were "effective" with respect to one set of outcomes (usually academic performance) were generally more "effective" in relation to others. This article reviews the last three decades of British evidence across a range of affective, social and other non-cognitive outcomes including: pupils' attitudes to school; their levels of involvement; truancy and attendance; citizenship issues; use of alcohol and drugs. The review draws two main conclusions. First, that schools' effects on these forms of outcomes do not seem as significant as those relating to academic ones. Second, that the evidence for a single underlying dimension of school effectiveness is limited. The factors which explain differences between schools in terms of academic progress appear to have limited explanatory power in relation to these "other" outcomes. The implications of this reassessment for the analysis of schools' functioning and the development of accountability systems are subsequently considered. (Contains 2 figures.)