Polysomnography (PSG) is the gold standard for the diagnosis of pediatric obstructive sleep apnea (OSA); however, high costs and limited availability restrict its use for routine screening. This systematic review and meta-analysis investigated the accuracy of overnight oximetry for the diagnosis of pediatric OSA. Studies evaluating overnight oximetry against PSG-derived apnea-hypopnea index (AHI) in subjects aged ≤18 years were considered in the qualitative analysis and evaluated with the QUADAS-2 tool. Only oximetry parameters adopted by at least four studies using the currently accepted diagnostic thresholds for pediatric OSA (AHI of 1, 5 and 10 events/h) were included for quantitative analyses. A bivariate meta-analysis was used to estimate sensitivity and specificity, as well as to construct summary receiver operator characteristic curves. The positive and negative predictive values were calculated. A total of 28 studies (9122 participants) were included in qualitative analyses. Only 3% oxygen desaturation index (ODI3) was eligible for the quantitative analyses (six studies, 1276 participants). As OSA severity increases sensitivity, specificity and the negative predictive value also increase, reaching values of 79%, 84%, and 89% at AHI ≥ 10, respectively. Oximetry displays a good performance as a screening tool for pediatric OSA, especially with moderate-to-severe disease. ODI3 is particularly effective at ruling out OSA in children who test negative.