This discussion focuses on some of the econometric issues that arise when one tries to estimate fertility equations on panel data sets of individual (micro) observations. As far as this author knows there has been only 1 published study (Heckman and Willis 1976) in which micro panel fertility equations were formally estimated. Most fertility equations estimated in other studies have used completed fertility rather than the number of children at each point in time as the dependent variable have used cross section rather than panel data and/or have used grouped rather than micro data. This paper presents estimates of micro panel fertility questions. The dynamic content of panel data directs attention to the timing of fertility in contrast to the focus on completed fertility with for example a cross section of data for women past childbearing ages. A point of departure is to modify the standard static demand for completed fertility model into an intertemporal model by means of a life cycle model with complete certainty. In such a model a couple at the beginning of marriage considers several lifetime budgetconstraints: the usual income expenditure constraint requiring that money expenditures over the lifetime on all goods including children equal total lifetime resources including nonwage income and the discounted value of the earnings of the husband and wife; a production function constraint showing the amount of the wifes time that must be devoted to childbearing; a human capital constraint showing the effect of being out of the labor force on future wages; and the usual time constraint in which the wifes available time is exhausted by time in leisure in childbearing (and associated housework) and in market work. Given the knowledge of the prices potential wage rates and other variables in these constraints the couple would choose a lifetime profile of all goods that yield utility including the number of children at each time K1. To provide an empirical illustration of the estimation techniques several fertility equations were estimated with data from the National Longitudinal Survey of Young Women (NLS). Both a tobit equation for the number of children and a probit equation adjusted for the truncation of the error term from below (as a result of the fact that the number of children cannot fall) were shown to yield similar estimates.