Quality of life differences across areas can be measured by differences in “real wages”, where real wages are computed as nominal wages adjusted for the cost of living. Computing cost of living differences involves several important issues, including how housing prices should be measured. Previous researchers typically have used some combination of rental payments and homeowner housing values, but housing values are forward-looking and may not reflect current user costs. This paper examines differences in quality of life estimates for U.S. metropolitan areas using, alternatively, rents and housing values. We find that the two measures of quality of life are highly correlated. Value-based estimates, however, are considerably more dispersed than rent-based estimates, likely because of the recent housing bubble and because housing values often provide an imperfect measure of the present user cost of housing. Researchers should be cautious in using housing values to construct quality of life estimates.