AIM: This paper aims to investigate the extent to which mindfulness-based interventions serve to reduce anxiety in children and adolescents. A heterogeneous sample was used, including clinical and non-clinical population. METHODS: A literature search of controlled intervention studies published up to December 31, 2016, was carried out in PubMed, Lilacs, Cochrane, Embase, PsycInfo, Opengrey and Teseo databases. The effect size was calculated by Cohen's d. The Cochran Q statistic and the I RESULTS: 829 articles were identified, of which 18 were finally selected. Of these, only three had statistically significant effect sizes. The overall combined result obtained was .013, but it did not result statistically significant (CI95% [-.102, .128].) The Q statistic was statistically significant (Q [18] = 28.497, P = . 39) and the I CONCLUSIONS: This meta-analysis did not obtain statistically significant results that could provide conclusions. In general, the studies analysed are small, of low power and have a marked heterogeneity, which implies that the findings are provisional and need to be supported by more robust studies. Although it cannot be ruled out that mindfulness-based interventions are not effective in the infant-juvenile population, it is also possible that this effect could not be detected due to the limited number of available studies. Larger investigations are needed, with sufficient statistical power and designs that control the variables potentially moderating, to establish clear conclusions.