This paper explores the household-level underpinnings of the observed aggregate relationship between consumption and wealth. There are important potential gains from studying wealth effects using data on individual households given that economic theory predicts that the response of consumption to movements in asset prices will differ depending on the source of those movements and on who is affected. Using data from the Consumer Expenditure Survey, I find that the consumption of stockholders had a strong positive correlation with current and lagged changes in stock prices in the 1980s and 1990s, whereas the consumption of non-stockholders did not. That pattern implies that changes in wealth had direct effects on spending as opposed to merely predicting changes in consumption because they signaled changes in future income. However, augmenting the sample with more recent data considerably weakens the result. A similar analysis of housing wealth effects for the years 1994-2008 suggests no near-term link between growth in house prices and growth in homeowners’ consumption of nondurables apart from housing itself.