Some selected alternative functional health measures are examined using marginal and conditional models in order to suggest a suitable measure for identifying the risk factors of functional limitations in elderly population of the USA. We have employed GEE with different correlation structure and other alternative marginal and conditional techniques and compared the results in terms of the identification of potential risk factors. For the purpose of this study, we have used the HRS data for the period 1992-2004. The GEE estimates show that age, gender, marital status, body mass index and race are significantly associated with functional limitations in elderly population. Other marginal models such as the one based on Azzalini (1994) also provide similar results. The implications of the results are discussed in the paper.