The techniques of percutaneous endoscopic gastrostomy are classified as the pull and introducer methods. Peristomal infection is the most common procedure-related complication in the pull method. The objective of this study was to compare procedure-related complications between the two methods and to assess the outcome of long-term use of the introducer method. Between January 1999 and November 2009, 116 patients received percutaneous endoscopic gastrostomy at Seoul St. Mary's Hospital in Korea. Before June 2006, the pull method was used for all patients; since then, every patient had been treated using the introducer method. We compared outcomes and complications within 180days of gastrostomy placement for the two methods. The pull method was performed on 63 patients and the introducer method on 53 patients. The occurrence of peristomal infection within 30days was significantly lower in the introducer method group than in the pull method group (1.9% vs. 36.5%, P=0.001). Methicillin-resistant Staphylococcus aureus, Pseudomonas aeruginosa, and Klebsiella pneumoniae were the major organisms isolated from the peristomal infection. The requirement for catheter reinsertion because of displacement, obstruction, or damage to the catheter between 30 and 180days was significantly higher in the introducer method group than in the pull method group (40.5% vs. 0%, P=0.001). Our results suggest the superiority of the introducer method in terms of infection control. However, the balloon-type catheter is associated with problems such as balloon insufficiency in long-term use.