BACKGROUND: Profound anatomical knowledge and surgical experience are essential for safe otological surgery. The surgeon's learning curve is evaluated in performing Gacek's singular neurectomy on cadaveric specimens. MATERIALS AND METHOD: One otological surgeon performed Gacek's approach on 96 halves of human heads embalmed according to Thiel's method, divided into four groups (24 halves per group) and evaluated them concurrent to the evaluation of an anatomist after a first surgical attempt. Successful operations were subdivided into "direct hits" of the osseous canal of the posterior ampullary nerve also known as the singular nerve and "indirect hits" with access to the posterior ampullary recess. Unsuccessful operations showed "no hit" of the nerve without lesion of the membranous labyrinth. "Indirect" or "no hits" were reinvestigated in a second attempt to evaluate possible reclassifications due to a learning process of the surgeon. The order of dissection, the rate of success and the changes of results in correlation with the numbers of dissected specimens were documented. RESULTS: The success rate significantly increased from 54.2% direct hits after the first group to 87.36% in the fourth group after the first attempt. Successful operations were performed in 86.5% after completion of the first attempt and 97.9% after the second attempt. The number of new allocations decreased from 11 cases in the first group of dissected specimens to zero in the fourth group. CONCLUSIONS: This paper strengthens the value of cadaveric training for surgeons and the crucial role of dissection of a large number of specimens in improvement of the surgeon's experience and success rate.