Background: Argon plasma coagulation is a diathermy-based non-contact therapeutic endoscopic modality that may have a lower risk of perforation than other tissue ablation techniques. Methods: Its effect was studied on three fresh esophageal and three fresh gastric resection specimens using power settings from 40 to 99 Watts at 90 degrees, with 1 mm separation using pulse durations of 1 and 3 seconds. A scoring system for depth of tissue damage was created and samples were analyzed blindly by a gastrointestinal histopathologist. Results: There was significantly greater damage to gastric tissue using a 3-second (compared with 1-second) pulse (p = 0.003) and marginally significantly greater damage to esophageal tissue using the 3-second pulse (p = 0.053). Tissue damage was related to power setting for gastric (p = 0.031) but not for esophageal tissue (p = 0.065). Only 1 of 42 esophageal samples and 2 of 42 gastric samples examined showed damage extending into the muscularis propria. Conclusions: Deep tissue damage that could lead to perforation was rare with argon plasma coagulation. The depth of gastric mucosal damage increased with increased pulse duration and increasing power settings, and, although the depth of esophageal mucosal damage was marginally related to pulse duration, it was not related to the power setting. (Gastrointest Endosc 2000;52:342–5).