Labyrinthine fistula is one of the most common complications of the chronic otitis media with cholesteatoma. Preoperative diagnosis of the labyrinthine fistula in the middle ear surgery is important for the preservation of the inner ear function. Opening of the bony labyrinth ofter creates inner ear damage through the disturbance of the inner ear environment and an accidental infection. In the retrospective charts review of the 1992 cases of chronic otitis media operated in Guro Hospital from 1984 to 1991, cholesteatoma was found in 544 patients (27%). The number of the labyrinthine fistula in the chronic otitis media with cholesteatoma was 32 (5.9%). Most of the fistula patients had had chronic otitis media for more than 10 years and the two age groups, males in their 20s and 30s were the most common. Preoperative diagnostic parameters showed generally low accuracy rate except the high resolution temporal bone CT. Half of the fistula patients had dizziness and 37% of patients showed spontaneous nystagmus preoperatively. The fistula test by Siegle or Politzer bag was positive in 29%. The temporal bone CT was performed in 18 patients preoperatively, and all 18 patients were diagnosed the fistula in the lateral semicircular canal, however, the fistula in the superior and posterior semicircular canal were not detected preoperatively. Most of the labyrinthine fistula patients were operated on by open cavity mastoidectomy : 23 patients (72%) were operated on with open cavity mastoidectomy, 6 (19%) radical mastoidectomy and, 3 (9%) intact canal wall mastoidectomy, respectively. The labyrinthine fistula was limited to lateral semicircular canal in 25 patients (78%), and 6 patients were showed the fistula in two more locations. The matrix in the fistula was removed in almost all except one patient. Facial nerve dehiscence was common in the labyrinthine fistula patients. A total 54% of patients had facial nerve dehiscence : 70% in the tympanic and 30% in the vertical portion. Postoperative hearing was evaluated in 24 patients. Ten patients were preoperatively deaf and fourteen patients had residual hearing. Among the fourteen residual hearer, six patients, had not had changed hearing more than 10 dB postoperatively and eight changed more than 10 dB postoperatively. One (8%) patient was became deaf and five improved hearing and two were worsen hearing slightly.