The body mass index (BMI) has become a universal index of energy nutritional status in adults even though it is influenced by many factors other than energy stores. One of these is variation in height caused by variation in the length of the trunk and legs. Such variations occur between and within populations. In the British population, with a sitting height/stature (SH@S) ratio at the mode of population values, adjustment of BMI for SH/S ratio caused BMI to change by 1 kg/m2 or more in 33% of women and 10% of men and by 2 kg/m2 in 5% of women and 1% of men. The possible effects of shape, as evinced by SH/S ratios should be borne in mind when describing individuals or populations as overweight or obese in epidemiological studies and in decisions of clinical management.