R ECENT attempts to measure changes in the quality of labor inputs in production have been based either on changes in years of schooling and other quality determining factors (Denison (1967) and Jorgenson and Griliches (1967)) or on changes in occupational mix (Griliches (1967) and Raimon and Stoikov (1967)).' Both measures raise serious problems. The first, because of lack of data, has not been able to adjust for differences in ability, on-the-job training or a number of other quality determining factors so that they do not even purport to have exhausted all significant quality differences.2 On the other hand, studies using occupational data have been based essentially on Bureau of Census major occupational groups i.e., eleven groups -and as the authors themselves were aware, these are uncomfortably broad. The purpose of this paper is to discuss still another (albeit related) approach to measuring changes in the quality of labor in production and to present the results for manufacturing industries. The approach is to first develop a wage rate index which measures changes in prices of a homogeneous bundle of labor services over time. The wage rate index is then used to deflate a time series of average hourly earnings (i.e., the cost of a changing bundle of labor services per man-hour) in order to obtain the desired index of quality per manhour. This index, in effect, measures changes in labor quality by changes in occupational mix -i.e., increases in quality per man-hour are measured by shifts from lower to higher paid occupations. The major difference between the approach used here and that in the previously cited studies is that my results are based on very detailed occupational data by sex and industry. Because of limited coverage, the indexes of quality per man-hour are for production (blue collar) workers in 25 selected manufacturing industries. On the basis of these series and some additional assumptions, I also attempt to expand the results to all employees in manufacturing in order to estimate the contribution of changes in labor quality to the growth in output and output per man-hour in this sector. In particular, changes in quality per manhour are divided between those resulting from shifts among production and nonproduction (white collar) workers and those resulting from changes in quality per man-hour within these two groups.