Background: Our understanding of the incidence of joint replacement across different subgroups of the Australian population is limited. This study investigated whether rates of hip and knee joint replacement vary according to socio-economic status, geographical locality, birthplace and indigenous status. Methods: Data from the National Hospital Morbidity Database were obtained. All separations for primary hip or knee joint replacement with the principal diagnosis of arthritis were identified. Age-standardized procedure rates were compared based on socio-economic status, geographical locality, birthplace and indigenous status. Results: Compared with age- and gender-matched individuals, rates of joint replacement overall were significantly lower in people living in the most disadvantaged (P < 0.05) and remote areas (P < 0.001), people born outside Australia (P < 0.05) and indigenous people (P < 0.001). However, there were some exceptions. Knee replacement rates were higher in the most disadvantaged areas than in the least disadvantaged areas (P < 0.01), and both hip and knee replacement rates were higher in regional areas than in major cities (P < 0.05). Males and females born in New Zealand and females born in the United Kingdom, Ireland and parts of Europe had a greater rate of hip replacement, and females from North Africa and the Middle East had a greater rate of knee replacement, than Australia-born people (P < 0.001). Conclusions: There is significant variation in the rates of primary total hip and knee replacement in Australia. Further investigation is needed to determine whether this simply reflects patterns of disease and disease risk factors, or is because of inequalities in health care, unmet need, and/or surgical or patient-related factors.