This monograph brings together current theory and research on atypical patterns of attachment in infancy and early childhood to illustrate and help in understanding some of the key issues in cases that do not "fit" the traditional attachment coding system developed by Ainsworth and her colleagues (Ainsworth, Blehar, Waters, & Wall, 1978). Researchers in this area have taken different perspectives in exploring mother-child relationships that are believed to be at special risk for abnormality and/or pathology, including variants of the disorganized classification and alternative forms of organized classifications. Their findings, however, are consistent in suggesting that as caregiving, the caregiving context, and/or infant functioning becomes more extreme, constraints on behavioral and relationship patterns implicit in the traditional categories of attachment may not operate. The extent to which discrete behaviors versus behavioral patterns, versus broader behavioral systems are useful for understanding normal and abnormal development also must come into question. These issues are addressed in conjunction with empirical findings in order to open up discussion on the topic and shed light on fruitful directions for future research. The volume opens with an introduction (Chapter 1) to attachment classification and the issues arising from identification of various forms of "atypical" attachments. Potential confounds between behaviors taken to be reflective of atypical relationships and symptoms of neurological impairments arising from congenital disorders (e.g., Down syndrome, cerebral palsy, autism) are discussed in Chapter 2. These two chapters set the context for a series of empirical studies that examine child cognitive functioning in a Down syndrome sample (Chapter 3), mother-child interaction (Chapter 4), infant emotional reactivity (Chapter 5), and maternal perceptions of child, self, and family as correlates of change in patterns across infancy (Chapter 6). Chapter 7 offers a developmental perspective on attachment organization in the context of developmental risk, and the final chapter summarizes and integrates the work presented in this "Monograph." Together, these chapters indicate the current status of theory and research about attachments that do not fit the traditional patterns among infants and young children at developmental risk. They also illuminate a variety of conceptual issues that warrant focused empirical attention in the next generation of research on parent-child attachment. Perhaps most important, they help strengthen the bridge attachment theory has forged, through its focus on pathways toward competence and pathology, between the fields of developmental and clinical psychology.