Purpose: To systematically review the literature to compare the bond strength of universal adhesives with etch-and-rinse and self-etch systems to primary teeth. Methods: The search was carried out in PubMed® /MEDLINE, Scopus® , LILACS, Embase® , and Web of Science™ databases with no restrictions. Two reviewers independently selected the studies, extracted the data, and assessed the risk of bias. Direct comparisons among universal adhesive in etch-and-rinse (UER) and self-etch (USE) modes and etch-and-rinse (ER) and self-etch (SE) systems were performed considering different substrates (sound enamel and dentin, and carious dentin) through meta-analyses of random effects. A mixed treatment comparisons meta-analysis was also performed comparing the bond strength of all adhesive approaches on sound dentin. Results: From 3,276 potentially eligible studies, 18 were selected for full-text analysis, and eight were included in the systematic review. All studies included in the meta-analyses evaluated a mild universal adhesive (Scotchbond™ Universal). In direct comparisons, there was no difference between USE and SE to sound enamel (mean difference [MD] equals 5.22; 95 percent confidence interval [95% CI] equals -9.09 to 19.52). In carious dentin, the results favored only ER over USE (MD equals -3.88; 95% CI equals -7.40 to -0.37). In sound dentin, the bond strength values of UER were higher than ER (MD equals 5.50; 95% CI equals 4.03 to 6.96). The rank probability showed that the best treatment on sound dentin was UER. Conclusion: Pooled in vitro data suggest that a mild universal adhesive system can substitute the etch-and-rinse and self-etch systems for restoring primary teeth.