Backgrounds/Aims: There is a limited study about the training program of unsedated transnasal esophagogastroduodenoscopy (UT-EGD) for a trainee who has no previous experience of this technique. The goal of this study was to determine the adequate level of training for technical competence in diagnostic UT-EGD for the trainee. Methods: From May 2008 to November 2008, three hundred patients who underwent UT-EGD under local anesthesia at a tertiary care academic medical center were included in this study. Four trainees who had no experience performing UT-EGD but very competent in conventional EGD, performed UT-EGD on 50 patients each. In a control group, four experts performed UT-EGD on a total of 100 patients. All trainees and experts used 5.2mm caliber endoscopes (GIF-260N, Olympus Medical Systems Corp., Tokyo, Japan). Technical success, procedure duration (intubation time to esophagus, intubation time to duodenal 2nd portion), complications, frequency of belching, and vital signs were checked. Results: Technical successes were 85%, 90%, 92.5%, 90%, 100% for every 10 patients totalling 50 patients each per trainee and 98% consistency in the control group. Intubation time to esophagus(sec) was 55.1 ± 26.9, 42.0 ± 13.6, 38.3 ± 13.1, 31.1 ± 7.4, 29.8 ± 6.6 in trainees and 31.5 ± 11.6 in the control group. Intubation time to duodenal 2nd portion(sec) was 130 ± 45.8, 121 ± 34.6, 110.5 ± 29.9, 102.5 ± 28.2, 98.5 ± 25.7 in trainees and 100.9 ± 27.4 in the control group, and showed no statistical significant difference after the first 3 lots of 10 patients. Epistaxis showed no statistical differences in both groups (18/200 vs 7/100), but the patients who complained of nasal pain were lower in the control group (44//200 vs 15/100). Frequencies of belching were 2.45 ± 1.36, 2.24 ± 2.06, 1.89 ± 1.72, 1.68 ± 1.99, 1.05 ± 1.27 in trainees and 1.39 ± 1.97 in the control group, and showed no significant difference after the first 3 lots of 10 patients. The willing to retry (164/200 vs 77/100) showed no significant differences in both groups, but patients who had previously experienced sedated EGD did not want to retry UT-EGD (26/186 vs 33/114). There were no significant difference in change of systolic and diastolic pressure, pulse rate, and oxygen saturation during UT-EGD between trainees and the control group, and there were no patients who experienced unstable vital signs and demanded oxygenation. Conclusions: Endoscopists who are competent in conventional EGD are able to obtain acceptable results from their first attempts, but 30 cases of experience is needed to acquire skillful hands for performing UT-EGD. Key words: Unsedated transnasal endoscopy, Learning curve