Loss of height occurs in the elderly. Not only is this height valuable to assess, but it creates difficulty for comparisons using equations based on estimates of stature in adult populations which often overlook the loss of height with age. Alternatives, such as the use of arm-span or hip length as surrogates for maximum stature (MS) during adulthood, have been proposed. In a study of 247 (130 men and 117 women) adult ethnic Chinese living in Melbourne, Australia, we tested the hypothesis that knee height is independent of age and attempted to devise an equation for the estimation of maximum stature (MS) in the this elderly group (aged 65 yrs) of this population. Anthropometric indices, including body weight, stature, arm-span, and knee height were twice measured using standard methods described by Lohman et al., and averaged for use in the analysis. In both men and women, the younger adults were taller and had a greater arm-span than their elderly counterparts: however, there was no difference in knee height or body weight between the two groups. Knee height was not associated with age, while stature and arm-span correlated negatively with age. These findings suggest that knee height provides for a valid estimate of MS during early adulthood than arm-span. Knee height is independent of age and does not appear to decreased over time, in spite of an expected cohort effect in this population. Arm-span, however, appears to change with a cohort as well as with age. Thus, there is a place in a life-time nutritional assessment of the aged to measure both arm-span as an index of cohort status and knee height for an individual's maximum achieved stature.