Over 80 percent of households in their 50s are homeowners and housing wealth accounts for over half of total household wealth for most of these homeowners. The evidence in the literature on whether the elderly are consuming their housing wealth has been mixed. Because home sales are infrequent and a high proportion of the elderly continue to own in old age, it appears that the elderly are not consuming housing wealth. There are, however, indications that housing wealth may be a form of self-insurance and that housing wealth is consumed, albeit at very old ages. To date, however, the evidence to support that hypothesis has been weak. This paper examines whether predictors of housing sales are consistent with the insurance story by looking at the extent to which indicators of changes in economic status and access to alternate insurance explain housing sales. The paper also examines the extent to which changes in health status predict housing sales. The results of the probit appear to indicate that, by and large, housing sales in old age for single households is mostly driven by worsening health. Widowhood has a large effect on increasing the probability of selling the house and the effect is larger if the husband is the surviving spouse. There are indications that poor married homeowners are consuming housing wealth and also indications that married households are responding to Medicaid tax incentives. This evidence seems to suggest that, at least among married households, housing decisions are financially motivated; however, the evidence does not by itself validate the insurance story.