Abstract Objective To evaluate the surgical results and complications in a cohort of patients operated on to repair a tegmen bony defect using either transmastoid approach with obliteration using S53P4 bioactive glass granules (TMA‐O), or the classic middle cranial fossa approach (MCFA). Study design A retrospective monocentric study. Methods Twenty‐five cases (24 patients) were included. Data regarding patient demographics, etiology, intraoperative findings, complications, recurrences, audiometric data, and follow‐up were analyzed. Results Seven patients were operated with MCFA and 17 patients with TMA‐O. One patient was operated on using a combined approach (MCFA + TMA‐O). In the preoperative HRCT scan, the size of the defect was estimated to be 6 ± 3.8 mm in the TMA‐O group and 6 ± 3.5 mm in the MCFA group ( P = .969). Intraoperatively, in the MCFA group, the location of the defect was mostly anterior (86%) with an intact ossicular chain (86%). A discontinuous chain was observed in 15 patients (88%) in the TMA‐O group. The mean follow‐up time was 22 ± 14 months in the TMA‐O group and 24 ± 15 months in the MCFA group ( P = .762). In both groups, there were no early postoperative complications or recurrences during follow‐up. Conclusion Repair of a tegmen bony defect with S53P4 bioactive glass granules seems to be safe and effective, limiting the use of the middle cranial fossa approach to cases with epitympanic defects and with an intact ossicular chain, and it could be used whatever the size of the defect and/or the presence of meningoencephalocele. Level of evidence 4.