The poor state of Aboriginal and Torres Strait Islander health has been documented in many ways, most obviously by comparing the relatively higher age-specific mortality and morbidity rates. This paper demonstrates the use of acute hospital separation data as a way to identify potential deficiencies in providing appropriate primary health care services for Aboriginal and Torres Strait Islander populations. It does so by using 'ambulatory sensitive conditions': those conditions (and procedures) for which high-quality appropriate primary health services deliverable under ideal circumstances are though to potentially reduce or eliminate the need for hospitalisation. Potential or realised access to primary care is not analysed directly using primary health service data. In this study, 1993-94 acute hospital separation data from NSW, Queensland, South Australia, Western Australia and the Northern Territory were used to calculate separation rates and odds ratios for Aboriginal and Torres Strait Islander and non-Aboriginal and Torres Strait Islander populations. Age-specific acute hospital separation rates for ambulatory sensitive conditions were 1.7 to 11 times higher for the Aboriginal and Torres Strait Islander populations studied. This supports clinical contentions that much Aboriginal and Torres Strait Islander morbidity and mortality is preventable and that further consideration is needed to service delivery reform at all levels in the health system and the distribution of funding.