INTRODUCTION: The purpose of this retrospective study was to provide basic probabilistic predictors of mortality to assist in determining appropriate therapeutic aggression in elderly burns population. METHOD: Eighty patients over the age of 70 years were admitted to the Victorian Adult Burns Service in Melbourne, Australia, over a period of 4 years. Retrospective data was analysed, taking into account patient demographics, type, site, depth and area of burn, presence of inhalation injury, number of co-morbidities, survival time and the number of operations performed, withdrawal of care and implementation of comfort measures only. RESULTS: Comparing survivors and non-survivors, significant differences were found between age, percentage total burn surface area (TBSA%), percentage full thickness surface area (FTSA%), presence of inhalation injury, site of burn and number of operations. The number of co-morbidities and gender were not significant to outcome. FTSA%, presence of inhalation injury, site of burn, age and number of operations were all significantly related to survival time. When patients who obtained comfort care were excluded from analysis, age and the number of operations were not considered to be significantly related to mortality. CONCLUSION: This study indicates that TBSA%, FTSA%, inhalation injury and age are significant predictors of death in the elderly burns population, although only the first three remain significant when patients who receive comfort care measures only are excluded.