There is at least one very perplexing characteristic of attachment in children: Why do children attach to abusive caregivers? Animal research provides clues to the answer to this question. Pain-related attachment is not unique to humans and has been observed in numerous species, including avian species and a myriad of mammalian species. Indeed, Bowlby, the father of Attachment Theory,18 built his model on the combined assessment of clinical work and animal research. First, the newly documented imprinting in birds suggested that attachment to the caregiver is innate or biologically determined. At hatching, chicks quickly learn to “attach” to or “imprint” on the first moving object they see, typically the caregiver, although a human or other animate object can be substituted. This imprinting occurs even when approach to the caregiver is associated with pain. Similar abusive attachment has been demonstrated in nonhuman primate colonies as well as other mammals such as infant dogs and rat pups. For example, shocking chicks during imprinting to the mother supports approach learning, while shock supports avoidance learning just hours after the imprinting sensitive period closes. Similarly, shocking or mistreating an infant dog while interacting with a caregiver still results in a strong attachment to the caregiver. This paradoxical attachment learning has also been demonstrated in nonhuman primates, including Harlow’s monkeys and more recently in other primate colonies, when abused infant monkeys form strong attachments to an abusive caregiver.19 Furthermore, children tolerate considerable abuse while remaining strongly attached to an abusive caretaker. It appears that selection pressure and evolution have produced an attachment system that ensures the infant attaches to the caregiver regardless of the quality of caregiving received.