Building upon previous evidence for the intergenerational transmission of antisocial behaviors, this research assessed and compared three models seeking to explain links between fathers' antisocial behaviors and children's behavior problems. A representative sample of children from low-income families (N=261) was followed from age 3 through age 9. Lagged OLS regression models assessed both short-term (1½ years) and longer-term (5½ years) prospective links between fathers' antisocial behaviors and children's behavior problems. Results supported a direct effects model: fathers' antisocial behaviors predicted growth in children's externalizing and internalizing behavior problems, with links stronger among resident-father families. Limited evidence of indirect effects emerged, with links between fathers' antisocial behaviors and children's behavior problems only slightly attenuated controlling for related risk factors and for parenting quality, showing limited evidence of mediation. A new interactive model was proposed and supported, with high levels of harsh discipline exacerbating negative links between fathers' antisocial behaviors and children's internalizing problems. Results suggest caution in policies and programs which seek to universally increase marriage or father involvement without attention to fathers' behaviors.