In the study of the determination of family size it is natural to focus on variables that describe the position of parents at the time of family formation. In fact, however, we do not have perfect measures of all that is relevant at this time; and for some of the "true" variables that matter, better proxies may be found in variables that describe the families from which the husband and wife came. This study explores the relationship between the number of children that couples have and some variables describing the background of the husband. When a man had an affluent or educated father and came from a small family or from a non-farm background, he and his wife tend to have fewer children than other couples. This is not just because these couples were themselves more affluent or educated; even when the "current" variables are introduced into the analysis, a statistically significant effect of the background variables persists. The effect of the background variables may be due to taste differences or to differences in prices and opportunities not adequately measured by the "current" variables. Professor Easterlin hypothesized that those who grew up in affluent parental homes would tend to have fewer children than others because of the tastes for a high standard of living acquired there. (Author/JM)