Introduction: Postoperative surgical site infections represent one of the most frequent complications related to health care and can increase morbidity and mortality, in addition to increasing treatment costs. Objectives: This study evaluated surgical site infections incidence after dentoalveolar surgeries of retained third molars. The article presents the development of a surgical complexity score that can be used as a predictive method for surgical site infections risk in dentoalveolar surgeries. Methodology: In this cross-sectional observational study, the sample was obtained for convenience, from medical records of patients diagnosed with tooth retention and who underwent dentoalveolar surgery of third molars, from 2015 to 2020. Demographic variables of patients, characteristics of the surgery and information regarding the diagnosis of surgical site infections in the postoperative period were collected. Results: A total of 360 medical records were analyzed, with 59 surgical site infection cases diagnosed (16.4%). There was a statistical difference (p<0.05) in the bivariate analysis correlating infection with time of procedure, type of incision and performance of ostectomy and tooth sectioning. Logistic regression analysis to identify the probability of infection showed no significance for any of the individual variables, except when they are combined composing the surgical complexity score, which was then significantly higher in surgical site infections cases. Conclusions: The surgical complexity score developed can be a new predictive tool for the occurrence of surgical site infections in dentoalveolar surgeries, taking into account that the greater the complexity of surgery, the greater risk.