Purpose: To systematically evaluate implant success rates with different loading protocols. Materials and Methods: A search was conducted of electronic databases, including The Cochrane Oral Health Group's Trials Register, PubMed, SciSearch, Medline, and EMBASE, for all randomized controlled trials published between 1997 and 2011 to compare implant success rates among different loading methods. The quality of randomized controlled trials was critically appraised, and the data were extracted by two independent reviewers. Meta-analyses were conducted of the eligible randomized controlled trials. Results: A total of 26 randomized controlled trials met the criteria for meta-analysis. The quality of these articles was moderate. Eight trials compared immediate and early loading (relative risk [RR] = 0.90, 95% confidence interval [CI] 0.42-1.93, P = .79), 7 compared early with delayed loading (RR = 1.19, 95% CI 0.52-2.72, P = .69), and 11 compared immediate and delayed loading (RR = 1.19, 95% CI 0.52-2.72, P = .69). Conclusions: The limited evidence shows that there is no significant difference in implant success rates with different loading protocols. Schlagwörter: delayed loading, dental implants, early loading, immediate loading, meta-analysis