Globally, the prevalence of Type 2 diabetes is rising. The most affected populations are those that have undergone recent and rapid transition towards a Western lifestyle, characterised by energy-dense diets and physical inactivity.¶ Two major hypotheses have attempted to explain the variation in diabetes prevalence, both between and within populations, beyond the contributions made by adult lifestyle. The thrifty genotype hypothesis proposes that some populations are genetically well adapted to surviving in a subsistence environment, and are predisposed to develop diabetes when the dietary environment changes to one that is fat and carbohydrate rich. The programming hypothesis focuses on the developmental environment, particularly on prenatal and early postnatal conditions: nutritional deprivation in utero and early postnatal life, measured by low birthweight and disrupted child growth, is proposed to alter metabolism permanently so that risk of diabetes is increased with subsequent exposure to an energy-dense diet. Both hypotheses emphasise discord between adaptation (genetic or developmental) and current environment, and both now put forward insulin resistance as a likely mechanism for predisposition.¶ Diabetes contributes significantly to morbidity and mortality among Australia’s Indigenous population. Indigenous babies are more likely to be low birthweight, and typical patterns of child growth include periods of faltering and rapid catch-up. Although there have been numerous studies in other populations, the programming hypothesis has not previously been tested in an Australian Indigenous community. The framework of the programming hypothesis is thus expanded to consider exposure of whole populations to adverse prenatal and postnatal environments, and the influence this may have on diabetes prevalence.¶ The present study took place in Cherbourg, a large Aboriginal community in southeast Queensland with a high prevalence of diabetes. Study participants were adults with diagnosed diabetes and a random sample of adults who had never been diagnosed with diabetes. Data were collected on five current risk factors for diabetes (general and central obesity, blood pressure, age and family history), in addition to fasting blood glucose levels. A lifestyle survey was also conducted. Participants’ medical records detailing weight growth from birth to five years were analysed with regard to adult diabetes risk to determine whether childhood weight and rate of weight gain were associated with subsequent diabetes. Adult lifestyle factors were xiialso explored to determine whether variation in nutrition and physical activity was related to level of diabetes risk.¶ Approximately 20% of adults in Cherbourg have diagnosed diabetes. Prevalence may be as high as 38.5% in females and 42% in males if those who are high-risk (abnormal fasting glucose and three additional factors) are included. Among those over 40 years, total prevalence is estimated to be 51% for females and 59% for males.¶ Patterns of early childhood growth may contribute to risk of diabetes among adults. In particular, relatively rapid weight growth to five years is associated with both general and central obesity among adult women. This lends some qualified support to the programming hypothesis as catch-up growth has previously been incorporated into the model; however, although the most consistent association was found among those who gained weight more rapidly, it was also found that risk is increased among children who are heavier at any age.¶ No consistent associations were found between intrauterine growth retardation (as determined by lower than median birthweight and higher than median weight growth velocity to one and three months) and diabetes risk among women or men. A larger study sample with greater statistical power may have yielded less ambiguous results.¶ Among adults, levels of physical activity may be more important than nutritional intake in moderating diabetes risk, although features of diet, such as high intake of simple carbohydrates, may contribute to risk in the community overall, especially in the context of physical inactivity. A genetic component is not ruled out. Two additional areas which require further investigation include stress and high rates of infection, both of which are highly relevant to the study community, and may contribute to the insulin resistance syndrome.¶ Some accepted thresholds indicating increased diabetes risk may not be appropriate in this population. Given the relationship between waist circumference and other diabetes risk factors and the propensity for central fat deposition among women even with low body mass index (BMI), it is recommended that the threshold where BMI is considered a risk be lowered by 5kg/m2 for women, while no such recommendation is made for men.¶ There are a number of social barriers to better community health, including attitudes to exercise and obesity, patterns of alcohol and tobacco use and consumption of fresh foods. Some of these barriers are exacerbated by gender roles and expectations.¶