Objective: The present study aimed to evaluate the safety of the first wearable augmented reality assistance system (ARAS) specifically designed for pancreatic surgery and its impact on perioperative outcomes. Background: Pancreatic surgery remains highly complex and is associated with a high rate of perioperative complications. ARAS, as an intraoperative assistance system, has the potential to reduce these complications. Methods: This prospective, single-center study included 20 patients who underwent pancreatic surgery using ARAS. These patients were matched in a 1:3 ratio with 60 patients from our retrospective data who underwent standard pancreatic resection. Matching variables were selected based on factors associated with poor intraoperative outcomes. Results: A higher proportion of patients in the ARAS group were diagnosed with borderline resectable pancreatic cancer and received neoadjuvant chemotherapy (20.0% vs 6.7%, P = 0.085). Additionally, more patients in the ARAS group required arterial resection compared with the control group (15.0% vs 0.0%, P = 0.002). Nevertheless, the ARAS group had a significantly shorter operative time (246 vs 299 minutes, P = 0.004) and required significantly fewer intraoperative blood transfusions (0.0 ± 0.0 vs 0.5 ± 1.4 units, P = 0.014). None of the patients in the ARAS group had positive resection margins (0.0% vs 20.0%, P = 0.045). Furthermore, patients in the ARAS group experienced a significantly shorter hospital stay (13.8 ± 6.6 vs 17.9 ± 8.2 days, P = 0.046). Conclusions: ARAS is a safe and effective assistance system for pancreatic surgery, offering superior perioperative outcomes compared with standard procedures.