BACKGROUND: Identification of resections with high risk of intraoperative complications is critical in guiding case selection for minimally invasive liver surgery. Several Japanese and European difficulty scoring systems have been proposed for laparoscopic liver surgery. However, the applicability of these systems for robotic liver resections has not been fully investigated. This study considers the Southampton system and examines its validity when applied to robotic hepatectomies. METHODS: We undertook a retrospective review of 372 patients who underwent robotic hepatectomies for various indications between 2013 and 2022. Of these patients, 63 operations were classified as low risk, 91 as moderate risk, 198 as high risk and 20 as extremely high risk based on Southampton criteria. Patient outcomes were compared by utilizing an ANOVA of repeated measures. Data are presented as median (mean ± SD). RESULTS: The Southampton difficulty scoring system was a strong predictor of intraoperative variables including tumor size, operative duration, estimated blood loss (EBL), and incidence of major vs minor resection (all < .0001). In contrast, the Southampton system was a weaker predictor of postoperative outcomes including 30-day mortality ( = .15), length of stay ( = .13), and readmissions within 30 days ( = .38). CONCLUSION: The Southampton difficulty scoring system is a valid system for classifying robotic liver resections and is a strong predictor of intraoperative outcomes. However, the system was found to be a weaker predictor of postoperative outcomes. This finding may suggest the need for proposal of a new difficulty scoring system for robotic hepatectomies.