The Pima Indians of Arizona have the world's highest reported prevalence of diabetes [1,2]; after 35 years of age, 50 % of the population is reported to have diabetes [1]. We have found that at each age, both the prevalence and incidence of diabetes is higher than earlier in our studies, resulting in more people with diabetes at earlier ages. As the age of onset of diabetes decreases, more women of childbearing age have Type II (non-insulin-dependent) diabetes or are at risk of developing gestational diabetes, which results in larger babies at birth [3±5]. Obesity, in turn, predisposes the children of women with diabetes to developing diabetes themselves, thus resulting in a ªcross-generational vicious cycleº [6] whereby successive generations of women are at greater risk of already being obese and having diabetes at childbearing age than the preceding generation.In healthy, nondiabetic women, normal pregnancy is associated with insulin resistance, hyperinsulin- Diabetologia (1998) Summary Children of women who have diabetes during pregnancy are more likely to become obese by early adulthood than those of women with normal glucose tolerance during pregnancy. Obesity can result from either excess food intake, low levels of energy expenditure or both. In our study, we tested whether maternal diabetes status influences total energy expenditure (TEE by doubly labelled water), resting metabolic rate (RMR by ventilated hood) and physical activity level (PAL = TEE/RMR and assessed by activity questionnaire). Measurements were taken in 88 5-year-old Pima Indian children, 24 children of women with diabetes (2-h plasma glucose ³ 11.1 mmol/l) diagnosed before or during pregnancy and 64 children of women with normal glucose tolerance (2-h plasma glucose < 7.8 mmol/l during pregnancy and no prior history of abnormal glucose tolerance). Although birth weight was higher in children of diabetic than of nondiabetic women (mean ± SD; 3.8 ± 0.6 vs 3.5 ± 0.4 kg, p < 0.03), there were no differences in weight (26.4 ± 6.9 vs 24.2 ± 5.6 kg) or per cent body fat ( 18 O dilution; 33 ± 8 vs 31 ± 8 %) between the groups at 5 years of age. There was no difference in TEE (6508 ± 1109 vs 6175 ± 942 kJ/d) or in RMR (4674 ± 786 vs 4483 ± 603 kJ/d) expressed as absolute values or after adjustment for weight and sex (TEE) or fat-free mass, fat mass, and sex (RMR). Physical activity level was also similar between the groups (1.40 ± 0.12 vs 1.38 ± 0.12). These results suggest that maternal diabetes status does not influence energy expenditure in the children by 5 years of age. Thus the greater obesity seen at older ages in the children of women with diabetes could be due to excess energy intake. Alternatively, if energy expenditure does have a role in the aetiology of obesity in these children, perhaps it does so only in older children. [Diabetologia (1998