The aim of the study was to identify potential failure modes (FMs) having a high risk and to improve our current quality management (QM) program in Collaborative Ocular Melanoma Study (COMS) ocular brachytherapy by undertaking a failure modes and effects analysis (FMEA) and a fault tree analysis (FTA).Process mapping and FMEA were performed for COMS ocular brachytherapy. For all FMs identified in FMEA, risk priority numbers (RPNs) were determined by assigning and multiplying occurrence, severity, and lack of detectability values, each ranging from 1 to 10. FTA was performed for the major process that had the highest ranked FM.Twelve major processes, 121 sub-process steps, 188 potential FMs, and 209 possible causes were identified. For 188 FMs, RPN scores ranged from 1.0 to 236.1. The plaque assembly process had the highest ranked FM. The majority of FMs were attributable to human failure (85.6%), and medical physicist-related failures were the most numerous (58.9% of all causes). After FMEA, additional QM methods were included for the top 10 FMs and 6 FMs with severity values > 9.0. As a result, for these 16 FMs and the 5 major processes involved, quality control steps were increased from 8 (50%) to 15 (93.8%), and major processes having quality assurance steps were increased from 2 to 4.To reduce high risk in current clinical practice, we proposed QM methods. They mainly include a check or verification of procedures/steps and the use of checklists for both ophthalmology and radiation oncology staff, and intraoperative ultrasound-guided plaque positioning for ophthalmology staff.