This study develops and tests two hypotheses regarding how childhood, adolescent, and postadolescent migration are associated with social integration in adulthood. Competing explanations are tested by estimating a number of models that control for earlier family context and contemporary adult characteristics. Using a national longitudinal and intergenerational data set, the results reveal that the age when a move occurred is associated with both higher and lower levels of social integration in adulthood. The direction of the associations depends on the sex of the offspring, the age at migration, and the measure of social integration. Adolescence appears to be the age when the effects of migration are most pronounced. Variables such as experiencing a parental divorce, low parental support, or growing up in a stepfamily and variables measuring adult characteristics of the adult offspring alter modestly the migrationintegration relationships. The process of international migration includes the extent to which migrants eventually integrate. The assumption is that the level of disruption associated with immigration affects one's ability to assimilate successfully, adapt, acculturate, and integrate following the move (Portes & Rumbaut, 1990). Yet, the stress and disruption accompanying migration are lower among migrants who can integrate quickly or who have access to social networks such as friends, relatives, or other ties in the destination community (Boyd, 1989). This concern with integration has not been a focus of studies on domestic migration-whether interstate, intercounty, or residential. That is, research on internal migration in the United States focuses on aspects other than postmigration integration, usually on human or financial capital outcomes, such as education, occupational mobility, and earnings. Recently, research has adopted a long-term orientation and an increased interest in the possible disruptive effects of earlier family migration on later adult outcomes perhaps, in part, due to the proliferation of longitudinal and intergenerational data and an emphasis on the entire life course. Yet, all of the recent longitudinal research on U.S. migration is along traditional foci-where migration is found to disrupt educational and occupational trajectories (e.g., Hagan, MacMillian, & Wheaton, 1996). A common perception is that migration during childhood and adolescence is disruptive and detrimental at the time of the move, shortly after the move, and, therefore, for social integration in adulthood. However, childhood and adolescent migration do not necessarily have to represent a stressful transition with negative consequences. Certain families may use migration as a way to move out of poor neighborhoods and school districts and to move into resource-rich neighborhoods (Parke & Bhavnagri, 1989; South & Crowder, 1997) with positive consequences for social and cognitive development. These differential effects of migration suggest two possible hypotheses. One, migration detracts from social integration. Two, migration enhances social integration. This research bridges the traditional concerns of international migration with the contemporary orientation of research on U.S. migration by examining the over-time association between childhood and adolescent migration and social integration in adulthood. BACKGROUND Social Integration The positive consequences of social integration are well documented. Higher levels of social integration in adulthood promote higher levels of social, psychological, and physical well-being. (See House, Umberson, & Landis, 1989, for a review.) Social integration, in terms of social support, is important as a resource to be tapped during stressful life events or transitional periods (Thoits, 1995). A general distinction often is made between structural and social-psychological or emotional integration (Moen, Dempster-McClain, & Williams, 1989). …