Introduction & Aims: In NOTES procedures it is essential to be able to perform secure closure of the access perforation. The aim of this study was to compare endoscopicaly sutured closure of a gastric access perforation using the T-tag tissue apposition system, (TAS) with closure via laparotomy in a randomized multicenter study. Methods: The gastric NOTES perforation was created using a needle knife and a 20 mm CRE balloon. For the study 30 pigs (18-38 kg) were used. After a diagnostic peritoneoscopy, where the pelvic organs were identified, the endoscope was withdrawn into the stomach. The animals were then randomized to either endoscopic closure with TAS-suture technique (in some cases with the addition of clips), or laparotomy with surgical closure using running resorbable 4-0 monofilament suture of the gastric perforation. The abdominal wall was closed with a running resorbable monofilament in the fascia and running nonresorbable suture for the skin. The time from randomization to endoscopic closure and withdrawal of the endoscope and operation time skin-to-skin were measured, as was recovery time and weight gain at day 7 and 14. At necropsy, adhesions, abscesses or peritonitis were recorded. Results: Of the 30 pigs, 27 survived without complications. Closure time using endoscopic technique was significantly longer than with open surgery, 23, 1 min (12-43) versus 12, 5 min (9-17,6). On average 2 sutures were needed for endoscopic closure. The median incision length in the surgical group was 11 cm (9-14). Recovery time post op was the same in both groups, but there was a tendency in favor of the endoscopic group. There was no difference in post op weight gain. One pig in the endoscopic group died within 12 hours post procedure in septicemia, without signs of leakage from the gastric perforation. Another pig in the Endoscopic group was euthanized after 7 days due to rectal prolapse, without any sign of leakage from the gastric perforation at necropsy. In the laparotomy group one pig was euthanized after 7 days due to incisional hernia, but no sign of leakage was recorded. At necropsy on day 14 there was significantly more intra abdominal adhesions in the laparotomised group, mainly between stomach, liver and abdominal wall. In both groups there were some adhesions between the gastric perforation site and the omentum. Five of the 16 pigs in the surgical group had wound infections. Conclusion: A gastric NOTES access perforation can safely be closed endoscopicaly using TAS sutures + clips. Moreover, animals in the endoscopic group had significantly less adhesions at autopsy and a tendency to recover faster post operatively, compared with open surgery.