Efforts to capture before and after visitation trajectories of fathers in romantic Visiting Parent Unions (VPUs) are sensitive to sample composition, estimator, visitation measure and theoretical perspective. Using the Fragile Families and Child Wellbeing Study, we examine extensive, intensive and sleepover margins of nonresident father involvement to reconcile theoretical and empirical differences in previous studies. We use generalized estimating equations (GEE) to address the non-normal distribution of visitation measures. At the extensive and the intensive margin we find VPU visitation levels before and after breakups are similar to and sometimes higher than ex-cohabiter levels. VPU visitation levels are always lower at the sleepover margin relative to ex-cohabiters. Our results challenge previous studies linking race and VPU status, confirm most previous perturbation results at extensive and intensive margins, and contribute a new result at the sleepover margin. We find the Baby Father Hypothesis to be relevant and fruitful for subsequent research.