The City Council of Parma and allied partners launched a programme based on professionally guided physical exercise and nutritional education in all primary schools of the metropolitan area, with a total of 7,000 children involved, ranging from 6 to 10 years of age. One scientific end-point of the programme was the definition of the parameter(s) most associated with increased physical activity and potential modification of eating habits. To this purpose we studied in a cohort of 2,151 9-year-old children the associations among BMI, fat mass (FM), breakfast eating and the amount of physical activity, the most relevant variables that characterize lifestyle. There was a consistent significant inverse correlation between FM and physical activity and a significant correlation between FM and breakfast skipping. When sorted by BMI, an inverse significant correlation was found between FM and physical activity in boys, except in those underweight. In girls, a significant correlation was found in those of normal weight, but not in those overweight or obese. The number of sports practised was related to FM only in overweight and obese boys. Breakfast skipping was significantly correlated with FM only in underweight girls. Taken together, our data show that FM can be used to accurately evaluate physical activity and eating habits in children, and suggest that, in preventive health programmes, the fundamental parameter to pay attention to is the amount rather than the type of physical activity.