The proposed research will investigate the effects of divorce and (re)marriage on family resources and children's well-being. It will address three central questions: 1. Does divorce or (re)marriage directly affect children, or are associations between these events and child well-being due to pre- existing differences between families who experience them? 2. Does divorce or (re)marriage affect family resources available to children -- economic, parental, and kin-community? Or are associations between these events and family resources also due to pre-existing differences between families? 3. To what extent do effects of divorce or (re)marriage on family resources account for effects of these events on children? We will use data from the National Surveys of Families and Households, conducted in 1987-88 and 1992-93, to estimate models of divorce and (re)marriage effects on family resources and child well-being. The first survey includes interviews with a primary respondent and her/his spouse or cohabiting partner in 1987-88. Follow-up interviews are being conducted with the same primary respondent, her/his current spouse or cohabiting partner, former spouses who were married to the primary respondent at the first interview, and a randomly selected focal child who was age 5-18 in NSFH-1 and who will be age 10-23 at NSFH-2. As a foundation for estimating structural models of divorce or (re)marriage effects, we will conduct measurement analyses to determine the stability of measures across time, and the relative quality of measures based on reports by parents and children in NSFH-2. Our primary models will be based on the observed panel data, using lagged endogenous variables. We will compare estimates from these models with those produced by fixed-effects and selection models where appropriate and feasible.For each of the central questions listed above, we will investigate potential interactions of structural effects with child's sex and with race/ethnicity. We will also investigate the extent to which effects of divorce or (re)marriage depend on child's age at event, and whether effects change over time.