Natural orifice transluminal endoscopic surgery (NOTES) has gained widespread interest as a potentially alternative to laparoscopic surgery, but concerns over peritoneal contaminations are unsolved. The aim of our study was to assess the safety of transluminal surgery by investigating the intraperitoneal bacterial load and contamination during transgastric and transvaginal surgeries. Twelve female pigs underwent transgastric (n = 7; tubal ligation and oophorectomy) and transvaginal procedures (n = 5; cholecystectomy). All animals were sacrificed after 2 weeks. The procedures were performed by using a double-channel endoscope (GIF-2T160; Olympus, Tokyo, Japan) under general anesthesia. Peritoneal fluid sampling was taken immediately after entry into the abdomen, at the end of the surgical procedure and during the autopsy, and sent for microbiologic assessment. In the transgastric group, 6 animals completed the surgical procedures and survived. Three pigs experienced signs of postoperative peritonitis with abscesses and adhesions visible and Escherichia Coli isolated at autopsy. In the transvaginal group, a cholecystectomy was performed without technical problems in all animals. No signs of postoperative sepsis nor growth in the microbiologic samples were recorded. In conclusion, the transvaginal approach seemed to be safer and produced less intra-abdominal contamination and sepsis, compared to the transgastric approach. Although both transgastric tubal ligation and oophorectomy and transvaginal cholecystectomy were feasible using equipment and accessories currently available for conventional endoscopy, new procedure-specific instruments and equipment should be developed to allow the operator safer access into the peritoneum.