Father involvement in family life among unmarried couples varies along a number of dimensions, including age of the child, relationship with the mother, race, parental education, and a number of economic indicators. There is a good deal of evidence to support the “package deal” hypothesis, which conditions father involvement on his relationship to the mother of his child. An alternative is the “baby father” hypothesis, which originated in the Caribbean and has not been studied as widely in an American context. The “baby father” hypothesis is the idea that unmarried father who were never co-resident with their child and the child’s mother follow a different father involvement pattern following the end of a romantic relationship with the mother, than their formerly co-resident peers. Specifically, fathers who were never in co-resident relationships have to negotiate their involvement with their children’s mother even before a breakup occurs; however they may be in a position of comparative advantage in so doing. Using four waves of the Fragile Families Child Wellbeing Study we test the “baby father” hypothesis with hierarchical linear models (HLM). HLM or multi-level models (MLM) permit the examination of individual performance patterns and the identification of factors related to those patterns by distinguishing between within-individual performance patterns and between-individual differences in those patterns. Our examination of trajectories of father involvement by initial mother/father relationship status, and by race, supports the “baby father” hypothesis. Understanding cultural differences in father involvement can provide salient information to public policy initiatives.