About one-fifth of children involved in investigations for abuse or neglect are placed in foster care. Although some return to their families quickly, others may remain in foster care for years without permanent family relationships. In this article, Mark Testa, Kristen Woodruff, Roseana Bess, Jerry Milner, and Maria Woolverton examine the Permanency Innovations Initiative (PII), a federally funded effort that tested innovative programs designed to prevent children from experiencing long stays in foster care and to build evidence for strategies that can be brought to scale in child welfare. PII aimed to follow a four-phase model for selecting, implementing, and testing interventions, including "exploration and installation, initial implementation and formative evaluation, full implementation and summative evaluation, and replication and adaptation." The results of the initiative were not encouraging. Some sites were never able to move to the full implementation phase. Others had significant trouble with participation rates. Two sites that were able to experimentally evaluate a fully implemented intervention found no significant differences between the treatment and comparison groups in achieving stable and permanent homes for children, and a third site found that the experimental results actually favored the comparison group. The authors "principal finding" is that "none of the promising innovations tested in this initiative yielded meaningful improvements in… stable permanence when rigorously evaluated." Discussing the implications for child welfare programs in general, they raise a fundamental issue: Should such programs primarily deal with maltreatment only after it has occurred? Or should they also work to prevent maltreatment from happening in the first place through early, universal interventions that strengthen protective factors within families?