Since more than 70 years, poverty is scientifically investigated as a risk factor for the cognitive and educational development of children and adolescents. Hundreds of programmes have been implemented which aim at preventing and compensating these unequal developmental precondition. The purpose of this dissertation was to synthesize the high-quality evidence worldwide into one comprehensive meta-analysis. An extensive, multimodal literature search for published and unpublished articles was conducted. Studies were included if the authors investigated the effectiveness of a psychosocial or educational programme designed to prevent the negative educational consequences of poverty. Additionally, the sample had to consist of children and adolescents up to the age of 18 years who predominantly (at least 70% of the sample) were of low socio-economic status (low-SES). Furthermore, studies were only eligible if a minimum of 50 participants were randomly assigned to the intervention and the untreated control group. A total of 109 relevant studies published between 1967 and 2013 could be retrieved. Most of them originated in the United States. The publications provided data regarding 132 contrasts between intervention and control groups. The sample included more than 80 000 children and adolescents predominantly between 2 and 12 years of age with mainly African-American or Hispanic ethnic background. Overall, an unbiased and robust, small but significant mean effect on children’s educational development could be found (d = 0.31). Stable long-term effects (more than 12 months post intervention) could be observed if children were approached directly but not if approached indirectly via their parents. Several moderating variables could be identified and further detailed analyses regarding child outcome constructs were conducted. The limitations of the dissertation are discussed and recommendations for research and practice are given.