Classic cardiovascular risk factors, such as smoking, arterial hypertension and hypercholesterolaemia, cannot explain a substantial part of the geographic differences in cardiovascular mortality. Anthropometric and nutritional factors in early stages of life may contribute to adult cardiovascular disease. Therefore, this work examines certain anthropometric variables and diet among children aged 6-7 y, living in four Spanish cities with widely differing ischaemic heart disease (IHD) mortality. Cross-sectional anthropometric and dietary survey in four cities in Spain. A total of 1112 children (50.1% males, 49.9% females) attending public and private schools in Cadiz and Murcia, cities with a relatively high IHD mortality, and Madrid and Orense, cities with a relatively low IHD mortality. A standardized method was used to measure anthropometric variables, and a food-frequency questionnaire completed by subjects' mothers, to measure diet. Body mass index (BMI), overweight (BMI>17.6 kg/m(2)), obesity (BMI>20.1 kg/m(2)) and intake of food and nutrients. Children in the four cities showed a high prevalence of overweight (range across cities, 28.9-34.5%) and obesity (8.5-15.7%). They also had a moderately hypercaloric diet (range, 2078-2218 kcal/day), marked by an excessive intake of lipids (45.0-47.3% kcal), particularly saturated fats (16.6-16.9% kcal), proteins (17.0-17.3% kcal), sugars (20.0-21.9% kcal) and cholesterol (161.6-182.9 mg/1000 kcal/day), and a low intake of complex carbohydrates (17.5-18.1% kcal) and fibre (19.6-19.9 g/day). Compared with children in the two low-IHD-mortality cities, those in the two high-IHD-mortality cities had a greater BMI (mean difference, 0.61 kg/m(2); P=0.0001) and ponderal index (0.58 kg/m(3); P=0.0001) and a higher intake of energy (104 kcal/day; P=0.007), cholesterol (16.00 mg/1000 kcal/day; P=0.0001) and sodium (321 mg/day; P=0.0001). Inter-city differences in anthropometric variables remained after adjustment for birthweight. Intake of fats, especially saturated fats, and cholesterol should be reduced among Spanish children. It could contribute to a needed reduction of the high prevalence of overweight and obesity in children. If the differences in anthropometric variables and diet between children from the cities with high and low coronary mortality are maintained in future or continue into adulthood, this could contribute to consolidate or even increase the IHD mortality gradient across cities. The finding that differences in anthropometric variables are independent of birthweight suggests that the childhood, rather than intrauterine environment, is involved in the development of such differences. This study was partly funded by grants from the International Olive Oil Board (Consejo Oleícola Internacional), Comunidad Autónoma de Madrid, Fundación Pedro Barrié de la Maza, and Fundación Eugenio Rodríguez Pascual.