The issue of how many elderly are affected by catastrophic nursing home expenses is a major part of the debate over if and/or how to reform long-term-care financing. Currently, there is some discussion regarding the magnitude of this catastrophic event, referred to as "asset spend-down", among the elderly. National data suggest the magnitude is small, while state-specific studies indicate it is greater. In addition, the literature regarding asset spend-down has presented two different measures of its magnitude, further confusing the issue. These two measures, each based on different denominators, have often been presented without adequate explanation. In this study, the authors review both measures and analyze reasons for the differences observed across studies. Major reasons identified include the type of sample used, the mix of payor source at admission, the length of time covered by the data, data on payor source/Medicaid eligibility, and the ability to observe multiple nursing-home stays within the data. Using the measure based on the number of persons who are private pay at admission, these studies indicate that approximately one fourth will eventually deplete assets. The second measure, based on a count of Medicaid residents at a point in time, indicates approximately one third were private pay when admitted. Study results indicate that national studies have underestimated the extent of spend-down due to national-level data limitations, while state-specific studies inevitably refect the specific state data set available and circumstances particular to each state. More state studies and a better understanding of asset transfer are needed.