ABSTRACT Aims To determine the major causes of mortality in weka (Gallirallus australis), and to investigate associations between causes of mortality and captivity status, age, sex, decade of submission, and season. Methods Necropsy records were obtained from the Massey University School of Veterinary Science/Wildbase Pathology database (Palmerston North, NZ) for weka submitted between 1 January 1995 and 22 March 2022. Causes of mortality were classified into categories based on aetiology. Frequency of diagnosis was tested for association with region of submission, captivity status, age, sex, decade, and season of death. Results A total of 156 necropsy reports were included in this study, of which 96 (61%) were from wild weka, 57 (36.5%) were captive, and three (1.9%) were of an unspecified captivity status. Weka were submitted from 12 regions across New Zealand. There were 65 (41.7%) adults, 16 (10.3%) juveniles, and 75 (48.1%) weka of an undetermined age among the 156 submissions. Of the weka with a known sex, there was a similar distribution between sexes with 27 (17.3%) males and 29 (18.6%) females. A cause of death was determined in 132/156 (84.6%) cases, with 24/156 (15.4%) cases having an unknown diagnosis. The leading cause of mortality in weka was traumatic injury, which occurred in 65/156 (41.7%), followed by infectious and/or inflammatory diseases in 26/156 (16.7%), and degenerative and/or nutritional conditions affecting 20/156 (12.8%) cases. The distribution of the primary causes of death was found to be dependent on captivity status (p < 0.001). Traumatic and toxic causes of death were more frequent in wild than captive weka. The cause of death was also dependent on season (p < 0.001). There was a significant difference in cause of death between summer and all other seasons (spring p = 0.008; autumn p < 0.001; winter p < 0.001) and between autumn and winter (p = 0.008). Conclusion Trauma was identified as the most significant cause of mortality in the free-living weka necropsied. The inherent and uncertain submissions biases, and low case numbers over a long period of time, means that temporal patterns and the effect of captivity status on causes of mortality should be interpreted with caution.