To describe the methodology behind the new UK Dietary Recommendations for Energy. Large interindividual variation in discretionary activity prevents prediction of energy expenditure, [total energy expenditure (TEE)], as a function of physical activity levels (PALs), [e.g. TEE = PAL × basal metabolic rate (BMR)] with the previously assumed accuracy. An alternative simplified approach derives PAL values in terms of population median and values for those exhibiting less or more activity (25th and 75th centiles) from suitable reference doubly labelled water (DLW) data. DLW data sets were used to identify PAL values: for adults, from two large randomized studies of urban populations (n = 929), and for children (aged 1-3, 3 to less than 10 and 10-18 years), from all published studies reporting mean values (n = 170 studies, 2502 patients) from which PAL values were reported or could be derived. Values for TEE were then derived from anthropometric determinants of BMR and PAL, and prescriptive estimated average requirement, (EAR) values were calculated on the basis of healthy body weights: BMI = 22.5 kg/m for adults and reference growth data for children. Energy costs of tissue energy deposition were modelled as a simple 1% increase in PAL values for children within the three age groups. The model for calculation of the EAR is simplified and widely applicable in most population settings.