Who cares for children and who works for wages in the labor market are basic family decisions. The purpose of this proposal is to quantify the impact of the price of child care on the type of child care parents use and on the employment decision of the mother, controlling for other attributes of child care and employment, such as quality, convenience, and wages. The proposed research will refine previous models of child care for employed mothers and develop new models of the joint child care and employment choice. A serious problem plaguing previous studies of the effect of price on child care choices is selectivity bias, arising because price is only observed for the choices actually made by parents. In this study, we are able to more effectively control for selectivity bias because the survey data we use include information on the alternatives respondents did not choose as well as those they did. Two ways of estimating characteristics of alternatives are proposed: 1) using parents' perceptions of their alternatives and 2) using exogenous information on the availability and attributes of child care choices in the same community obtained through concurrent surveys with providers. The model will be used to conduct policy simulations of the effects on choice of changing the attributes of child care and employment and of changing the availability of care. The data come from three existing data sets linked together by county: the 1990 National Child Care Survey, the Profile of Child Care Settings Study, and a Contextual Data File. The impact of price and other attributes of care, characteristics of the family and characteristics of the area in which they live on choice of child care and employment will be examined using discrete choice models, including the mixed multinomial/ conditional logit model, the joint logit model, the nested logit model, and the universal logit model. As far as we know, this is the first attempt to formally link supply and demand for child care in U.S. communities.