The purpose of this study is to take a new look at factors affecting the spending of retired elderly households by identifying the effects of different types of wealth components, along with socio-demographic factors, on current consumption. A tobit linear regression model was utilized to estimate parameters in the consumption functions. Data was drawn from the 1990 Consumer Expenditure Survey, Interview Survey in the United States.Total consumption and various consumption categories were the most responsive to changes in annual income flow and the least responsive to changes in property assets. The four wealth components differed in their influence on consumption categories among retired elderly households. In addition, age, residential area, household type, and education were found to be significant factors affecting total consumption and consumption categories of the retired elderly.