Objective: Through comparing the short term efficacy of laparoscopy-assisted pancreaticoduodenectomy (LAPD) and open pancreaticoduodenectomy (OPD) to assess the feasibility of LAPD and its conformity to the principles of radical cancer surgery and minimal invasiveness. Methods: The clinical data of 79 patients undergoing pancreaticoduodenectomy (PD) from May 2012 to December 2014 were retrospective analyzed. Of the patients, 25 cases received LAPD (LAPD group) and 54 cases had open surgery (OPD group). The perioperative variables between the two groups were compared. Results: In LAPD group compared with OPD group, the operative time (474.6 min vs. 404.9 min), surgery cost (21 500 yuan vs. 18 700 yuan) were increased, while the intraoperative blood loss (265.5 mL vs. 380.8 mL), intraoperative transfusion rate (8.0% vs. 29.6%), times for postoperative analgesia demand (0.56 times vs. 1.6 times), time to first flatus passage (3.8 d vs. 5.7 d), time to first liquid intake (5.3 d vs. 7.6 d), length of ICU stay (0.74 d vs. 1.3 d), length of hospital stay (15.5 d vs. 19.0 d) and tumor size (2.0 cm vs. 3.9 cm) were all significantly reduced (all P<0.05). There was no significance between the two groups in incidence of overall or individual complications, mortality, R0 resection rate, the number of resected lymph nodes and the number of positive lymph nodes (all P>0.05). Conclusion: Although the procedure of LAPD is complicated and troublesome, it is safe, feasible and can follow the principles of radical cancer surgery and minimal invasiveness in some selected patients, for surgeons who have a full command of OPD technique and adequate laparoscopic experience after completion of certain learning curve.