The contribution of children to viral spread in schools is still debated. We conducted a systematic review and meta-analysis of studies to investigate SARS-CoV-2 transmission in the school setting. Literature searches on 15 May 2021 yielded a total of 1088 publications, including screening, contact tracing, and seroprevalence studies. MOOSE guidelines were followed, and data were analyzed using random-effects models. From screening studies involving more than 120,000 subjects, we estimated 0.31% (95% confidence interval (CI) 0.05–0.81) SARS-CoV-2 point prevalence in schools. Contact tracing studies, involving a total of 112,622 contacts of children and adults, showed that onward viral transmission was limited (2.54%, 95% CI 0.76–5.31). Young index cases were found to be 74% significantly less likely than adults to favor viral spread (odds ratio (OR) 0.26, 95% CI 0.11–0.63) and less susceptible to infection (OR 0.60; 95% CI 0.25–1.47). Lastly, from seroprevalence studies, with a total of 17,879 subjects involved, we estimated that children were 43% significantly less likely than adults to test positive for antibodies (OR 0.57, 95% CI 0.49–0.68). These findings may not applied to the Omicron phase, we further planned a randomized controlled trial to verify these results.