Although research on the consequences of adolescent parenting for the children of adolescent mothers has focused on these children's increased risk for maladaptive outcomes (see reviews in Brooks-Gunn & Furstenberg, 1986; Furstenberg, Brooks-Gunn, & Chase-Lansdale, 1989; Osofsky, Hann, & Peebles, 1993), not all children of adolescent mothers suffer negative consequences. In the extensive 17-year study of 296 children of adolescent mothers in Baltimore (Furstenberg, Brooks-Gunn, & Morgan, 1987), 53% did experience negative academic outcomes: They repeated at least one grade. However, 26% of the sample were above average in class standing; Furstenberg et al. (1987) documented that three changes in an adolescent mother's family life positively affected her child's academic status: terminating welfare status, contracting a stable marriage, and advancing her own education. Similar results were recently reported by Barratt (1991). Although the academic performance of the Baltimore sample of children of adolescent mothers was strongly affected by mothers' economic, educational, and marital status, children's behavior problems were only influenced weakly by these same variables (Furstenberg et al., 1987). The absence of strong effects suggests that other family characteristics need to be examined as predictors of behavior problems in children of adolescent mothers. One clear candidate for a predictor of the behavior problems of children of adolescent mothers is maternal parenting practices. Parenting practices were not examined in the original Baltimore study but have subsequently been found to contribute to infant behaviors (Field, Widmayer, Adler, & de Cubas, 1990) and have been proposed to be potentially important predictors of later outcomes as well (Furstenberg et al., 1989). However, this proposal has yet to be tested systematically. The continuing lack of longitudinal research on the impact of adolescent parenting practices on children after infancy has led to repeated requests for such research (Brooks-Gunn & Chase-Lansdale, in press; Chase-Lansdale, Brooks-Gunn, & Paikoff, 1991; Furstenberg et al., 1989; Roosa, 1991). The current study is the first of a series of studies at sites in Kansas, Louisiana, and Oklahoma designed to remedy the lack of such longitudinal research on adolescent mothers and their children by investigating the relationship between parenting practices of adolescent mothers from infancy through early childhood and their preschool children's behavior problems and positive socioemotional outcomes. This series of studies is grounded in a new conceptual framework of adolescent parenting that modifies attachment theory's emphasis on the primary importance of the mother-child relationship in predicting later child outcomes (Bowlby, 1969, 1973, 1980; Ainsworth, Blehar, Waters, & Wall, 1978; Main, Kaplan, & Cassidy, 1985; Sroufe, 1979, 1989) by adding the adolescent mother's own continuing development to the prediction of her child's later behavior. Thus, the first purpose of the current study is to remedy the lack of longitudinal research on the impact of parenting practices of adolescent mothers on their children and the second purpose is to test the proposed model of adolescent parenting. Attachment theory and research on attachment in adult mothers and their children emphasize the continuity from children's secure attachment in infancy to their social competence in childhood (Waters & Sroufe, 1983; Waters, Wippman, & Sroufe, 1979) and from children's insecure or disorganized attachment to their behavior problems in childhood (Greenberg, Speltz, Deklyen, & Endriga, 1991; Sroufe, 1989; Sroufe & Rutter, 1984). Some researchers argue that this continuity in children's adaptation may well be due to the stability of and low stress in the lives of their adult mothers (e.g., Lamb, 1987). Stability and low stress are not hallmarks of the lives of adolescent mothers. …