Back schools are educational programmes for the prevention and rehabilitation of back pain. A quantitative review (meta-analysis) was performed to synthesize the existing evidence on their effectiveness, for different outcome criteria and time categories. On the basis of a systematic literature research, 18 controlled back school studies with a total of 1682 participants could be included in the meta-analysis. Effect sizes that compared back school patients with patients in a control group were calculated for 14 outcome criteria and four time categories. Effectiveness of back schools was shown for the time period within 6 months of intervention. However, effects for the 14 examined criteria revealed large differences: in contrast to the strong effects on correct back posture and movements and on knowledge of back school contents, the intervention had only small effects on health economic variables (e.g. utilization of the health care system) and no effects on clinical variables (e.g. pain intensity). With regard to the time period following the 6 months after intervention, the analysed data does not strongly support the effectiveness of back schools. It is concluded that the effectiveness of back school intervention depends on the outcome criterion and time of measurement. The results suggest that the concept of back school programmes should be improved and systematically re-evaluated.