Social science research that can be translated into policy recommendations pertaining to the custody of children after divorce is meager. Giving specific attention to the debate on the advantages and disadvantages of joint custody, this article proposes multilevel-multivariable life cycle guidelines for future child custody research. Critical issues are discussed, empirical questions raised, and salient variables examined for both the divorced family and the social system. For the divorced family, six classes of variables (parent-child relationships, interparental relationships, mechanics of alternations, similarity-dissimilarity of home environments, characteristics of children, and social-demographic characteristics) are discussed. For the social system, attention is given to the potential effect of formal social institutions (work settings, schools, and the legal system) and of informal social networks (kin and friends) on the postdivorce interrelationships of parents and children. The importance of examining the effects of custody arrangements in terms of a family life cycle is emphasized. The remarriage of one or both parents is used to illustrate how the effects of joint custody may be altered by anticipated changes in the life cycle of divorced parents and their children.Each American child learns, early and in terror, that his whole security depends on that single set of parents who, more often than not, are arguing in the next room over some detail in their lives. A desperate demand upon the permanence and all-satisfyingness of monogamous marriage is set up in the cradle. What will happen to me if anything goes wrong, if Mommy dies, if Daddy dies, if Daddy leaves Mommy, or Mommy leaves Daddy? are questions no American child can escape.