This paper uses data from the baseline Fragile Families and Child Well-being Study to examine the level and effects of father-involvement on child's birth weight and mother's health behavior during pregnancy (prenatal care, drinking, drug use and smoking). The findings indicate that most fathers, including unwed fathers, are involved with their children at birth and have intentions to remain involved. The effects of father involvement on health and health behavior depend, however, on how the construct is measured. When measured as parent's relationship status (married, cohabiting, romantic or non-romantic), the effects of marriage are beneficial for all but one outcome, the effects of cohabitation are positive for prenatal care only, and the effects of romantic involvement are negative for child's birth weight. When measured as paternity acknowledgement, contributions during pregnancy and intentions to contribute, unmarried father involvement has no effect on child's birth weight, a strong effect on early prenatal care and a variable but overall positive effect on mother's health behaviors. Furthermore, the effects of father involvement do not vary systematically by fathers' earnings potential and psychosocial attributes. While these results support the notion that fathers can influence mothers to maintain or adopt healthy pregnancy behaviors, they do not indicate that father-involvement improves birth outcomes.