Although pancreaticoduodenectomy (PD) outcomes have improved, it remains a procedure with high perioperative complications. Surgical site infection (SSI) is one of the most common complications after PD, but difficulty remains in identifying high-risk patients. We present a risk stratification tool to predict patients at highest risk for SSI. Methods: Data was retrospectively collected on all patients undergoing PD at a tertiary hospital (9/2011e8/2014). Univariate analysis was performed to determine factors individually associated with SSI. These were incorporated into a multivariate logistic regression model with receiver operating characteristics analysis. The resulting odds ratios were converted into a point system to create a SSI score with internal validation. Results: 679 patients underwent PD and were split into test (443 patients) and validation groups (236 patients). There was no difference in patient or perioperative demographics between groups. Thirty-day SSI was observed in 16.7% (n = 114). By univariate analysis, history of chemoradiation, pre-operative chemotherapy and/or radiation, bile stent, absence of a superficial wound vacuum device, and vascular resection were associated with SSI (all p < 0.05). On multivariable analysis, pre-operative bile stent or drain and neoadjuvant chemotherapy were independent predictors of SSI (all p < 0.001). Predicted risk of SSI using this score was 0%, 32%, and 64% for 0, 1, and 2 points, respectively (AUC = 0.73, R 2 = 0.93). The model performed equivalently in the validation group (AUC = 0.77, R 2 = 0.99). Conclusion: This novel, validated risk stratification score accurately predicts SSI risk after pancreaticoduodenectomy. Identifying patients with the highest risk of SSI can help target interventions to reduce SSI, such as superficial wound vacuum devices.