Childhood depression is a serious mental problem and has detrimental effects on child development. The Children’s Depression Inventory (CDI) has been commonly used to measure childhood depression in developmental research across the world. The present study aimed at meta-analyzing estimates of Cronbach’s alpha reported in previous studies and identifying study characteristics that affect score reliability. The final sample included 331 alphas obtained from 114,023 children in 283 primary studies conducted in 44 countries and areas by using 25 different languages. Although alphas reported in the literature were generally high with a point estimate of .840, 95 % CI [.834, .846], considerable heterogeneity existed between studies. The long form of CDI with 27 items had higher alphas than the short form with 10 items. Sample size and group heterogeneity were correlated with alpha. Moreover, alphas did not differ across linguistic forms of the CDI, indicating the cross-cultural equivalence of score reliability. Implications for future applied and methodological research were discussed.