To evaluate the vaccine effectiveness (VE) of mRNA COVID-19 vaccines in children using a meta-analysis approach. Relevant studies on the use of mRNA COVID-19 vaccines in children were identified through computerized searches. VE-related indicators were extracted, and data analysis was performed using the R software with the meta-package. This study included a total of 12 relevant articles involving 9,963,732 participants from multiple centers in different countries, including the United States, Canada, Singapore, Israel, South Korea, and Qatar. The administered vaccine types included BNT162b2 and mRNA-1273. Participants were categorized into partially immunized (one dose of vaccine) and fully immunized (two doses of vaccine). Four articles reported VE after one dose of vaccine, while 12 reported VE after two doses. Heterogeneity analysis indicated significant heterogeneity among the studies, warranting the use of a random-effects model for analysis. Meta-analysis results revealed that the VE of partial immunization ranged from 16.61 (95% CI: 6.32-25.77) to 34.30 (95% CI: 24.21-43.04), with a pooled VE of 22.80 (95% CI: 15.68-29.32). The VE after full immunization ranged from 16.14 (95% CI: 14.42-17.83) to 90.47 (95% CI: 67.42-97.21), with a pooled VE of 56.17 (95% CI: 41.12-67.37). Meta-regression analysis showed no statistically significant correlation between VE and time (p>0.05). Both partial and full immunization of the BNT162b2 mRNA vaccine provide benefits in reducing infection rates. VE varies over time and is closely associated with viral mutations and waning immunity. The specific mechanisms require further investigation.