Labeling theorists have argued that members of disadvantaged groups are disproportionately likely to be labeled negatively but we find that in some situations members of disadvantaged groups may, in fact, be less likely than others to be labeled for the same actions. In neonatal intensive care units (NICUs), we argue, the process of sorting and evaluating parents of NICU patients varies with the gender age, and race of the parents and depends on what those characteristics are taken to mean in the context of the organization. We find that social control in the NICU tends to be reintegrative rather than disintegrative. Because NICUs cannot control their organizational borders and because of the fundamental interdependence between medical careproviders and parents, parents must be reformed rather than excluded We conclude that when substantial resources must be expended to correct deviant behavior, an organization will be more reluctant to label anyone as deviant and that when a group is dependent on the participation of deviant members, the group will make considerable effort to rehabilitate them.