By retrieving literature published from 2005 to 2015 from Chinese National Knowledge Infrastructure, Wanfang, Vip, PubMed, and Web of Science, we filtered out studies using the Children’s Depression Inventory only and compared left-behind children and non-left-behind children. The methodological quality of the papers was evaluated using the Newcastle–Ottawa Scale. Finally, we included six studies to carry out a meta-analysis. The results showed that the Children’s Depression Inventory scores of left-behind children are significantly higher than those of non-left-behind children (standardized mean difference: –0.233, 95% confidence interval: –0.036 to –0.430, p < 0.05).