This paper uses the data in the NBER/CPE pilot sample of genealogies to create a new time series on life expectation in the U.S. since 1720. After attaining remarkably high levels toward the end of the eighteenth century, life expectation as measured by ê began a decline that lasted about 80 years before beginning the new rise with which we have long been familiar. Second, time series on the average adult stature of national populations in North America and Europe are used as a measure of nutritional status. The properties of this measure in the analysis of labor welfare and an explanation for the high correlation between stature and the Cmi ratio are discussed. The time series on stature is strongly correlated with the series on and other measures of mortality. Third, these correlations are used to estimate the contributionof improvements in nutritional status (not diet alone but diet net of prior claims) to the decline in mortality in Europe and America since 1800. Improvements in nutritional status may have accounted for as ifiuch as four tenths of the decline in mortality rates, but nearly all of this effect was concentrated in the reduction of infant mortality. The new findings are used to resolve several paradoxes and the implication of the findings for the standard-ofliving controversy are considered.