Background Medication safety incidents commonly occur in mental health hospitals. There is a need to improve the understanding of the circumstances that are thought to have played a part in the origin of these incidents to design safer systems to improve patient safety. Aim This study aimed to undertake a mixed-methods analysis of medication safety incidents reported to the National Reporting and Learning System in England and Wales in 2010 to 2017. Method Quantitative analyses of anonymized medication safety incidents occurring in mental health hospitals that were reported to the National Reporting and Learning System during an 8-year period were undertaken to characterize their type, severity, and the medication(s) involved. Second, a content analysis of the free-text reports associated with all incidents of at least moderate harm severity was undertaken to identify the underlying contributory factors. Results Overall, 94,134 medication incident reports were examined, of which 10.4% (n = 9811) were reported to have resulted in harm. The 3 most frequent types of reported medication incidents involved omission of medication (17,302; 18.3%), wrong frequency (11,882; 12.6%), and wrong/unclear dose of medication (10,272; 10.9%). Medicines from the central nervous system (42,609; 71.0%), cardiovascular (4537; 7.6%), and endocrine (3669; 6.1%) medication classes were the most frequently involved with incidents. Failure to follow protocols (n = 93), lack of continuity of care (n = 92), patient behaviors (n = 62), and lack of stock (n = 51) were frequently reported as contributory factors. Conclusions Medication incidents pose an enduring threat to patient safety in mental health hospitals. This study has identified important targets that can guide the tailored development of remedial interventions.