Although a large number of studies have examined associations between paternal involvement and children's outcomes, most are based on a single source of data or fail to control for maternal involvement. We used data from the National Survey of Families and Households (n = 994) to test the hypothesis that positive father involvement is associated with fewer behavior problems in children. To avoid same-source bias, we used fathers ' reports of involvement with children and mothers' reports of children's behavior problems. To determine if fathers make a unique contribution to their children's behavior, we controlled for mothers' reports of maternal involvement. Structural equation models revealed that positive paternal and maternal involvement were independently and significantly associated with children's behavior problems. Estimated effects were similar for biological fathers, stepfathers, White fathers, Black fathers, and Latino fathers. Key Words: Black fathers fathers, fathers, Latino fathers, parent-child relations, stepfathers. We examine the notion that fathers' positive involvement with children in two-parent households-as reflected in shared activities, supportive behavior, and feelings of affection-has beneficial implications for children's behavior and development. Although many people assume that a positive father-child relationship benefits children, the evidence for this assumption is surprisingly weak. Many studies demonstrate statistically significant associations between paternal involvement and child outcomes. (See Biller, 1993; Lamb, 1997; Popenoe, 1996; and Snarey, 1994, for reviews.) Two major limitations, however, are common in this research. First, many studies use the same source of information on fathers' behavior and children's outcomes. For example, children may report how close they feel to their fathers, as well as their own self-esteem. Observed associations in these single-source studies are likely to be inflated by shared-method variance. Second, many studies fail to control for the quality of the mother-child relationship when estimating the impact of the father-child relationship. Mothers who are especially effective parents also may be supportive of paternal involvement with children. If this is true, then observed associations between paternal involvement and child outcomes may be spurious. To determine if fathers are uniquely important to children's well-being, it is necessary to include the quality of the mother-child relationship in analyses. The study presented here contributes new evidence of the links between paternal involvement and child outcomes using the 1987-1988 National Survey of Families and Households (NSFH). To avoid problems with shared-method variance, we use fathers' reports of involvement with children and mothers' reports of children's behavior. To see if paternal involvement is associated independently with children's behavior, we control for the level of maternal involvement. In addition, we consider whether estimated father effects vary between biological fathers and stepfathers and between fathers from different racial and ethnic groups. CONCEPTUAL FRAMEWORK As children develop, they actively use the resources in their social and physical environments. Because fathers and mothers are usually (but not always) the central members of children's social networks, children's development is bound closely with the quantity and quality of resources that parents provide. Perspectives in sociology (Coleman, 1988) and developmental psychology (Lamb, 1997) assume that fathers contribute to their children's development, not only through the provision of financial capital, but also through the quality of their interactions with children. For this reason, Coleman used the presence of the father in the household as a key indicator of children's social capital. Later studies, however, have extended Coleman's work to focus on specific paternal behaviors because fathers who are inattentive or abusive do not provide social capital to their children. …