ObjectiveTo determine the predictive factors for residual disease occurring after surgical removal of congenital cholesteatomas and whether these predictive factors differ between microscopic ear surgery (MES) using data from the literature and transcanal endoscopic ear surgery (TEES) using data from our own institution. MethodsTwenty-three patients with a congenital cholesteatoma who underwent surgical treatment at Yamagata University Hospital between December 2011 and December 2017 were retrospectively investigated. We divide TEES into three different approaches: non-powered TEES, powered TEES and dual MES/TEES. Main outcome measures were Potsic stage, closed or open congenital cholesteatoma type, TEES surgical approach, appearance of residual disease, tympanoplasty type and hearing outcome. ResultsA logistic regression analysis was conducted on the Potsic stage, closed or open type, TEES surgical approach and age to obtain the odds ratio for residual disease. The chance of residual disease significantly increased in the presence of an open-type congenital cholesteatoma (odds ratio: 30.82; 95 % confidence interval: 1.456–652.3; p = 0.0277), but not for any of the other factors including Potsic stage. The timing of the confirmation of residual disease after ossicular chain reconstruction was analyzed using a Kaplan-Meier analysis. The residual disease rate was significantly higher with an open-type congenital cholesteatoma (log-rank test, p < 0.05). In addition, all residual disease occurred within three years after surgery. ConclusionsOur results showed that an open-type congenital cholesteatoma is the strongest predictive factor for residual disease when removing a congenital cholesteatoma by TEES.