Attachment theory is a core theory of child development. The theory proposes a framework to understand how children’s early relationships to their caregiver shape children’s lifelong development. Most attachment research, however, is limited to categorical assessments of infant-mother attachment. This results in a reductionist understanding of children’s development, one that rests on a number of questionable assumptions. From an assessment perspective, categorical measures of attachment assume that all children fit neatly into a fixed number of categories, and that all children within a category present similar attachment behaviors. From a developmental perspective, a focus on infant attachment assumes little change in children’s caregiving environments, and this, despite evidence showing that changes may occur between infancy and the preschool years. Such changes influence child-caregiver attachment relationships. From a caregiver perspective, children’s relationships with their mothers are influential, but they do not span the gamut of children’s early relationships. Fathers, for example, are increasingly involved in child rearing and are influential in children’s development. This dissertation sought to overcome these limitations through the use of the Preschool Attachment Rating Scales (PARS), a novel, continuous measure of child-caregiver preschool attachment. The first study demonstrated the reliability and validity of the PARS as a measure of child-mother and child-father preschool attachment. This study examined the inter-rater reliability, the convergent validity, the construct validity, the predictive validity, and the incremental validity of the PARS. The second study used a longitudinal design to examine the independent and interactive influence of child-mother and child-father attachment in the preschool years on boys’ and girls’ externalizing behaviors in middle childhood. This latter study showed that the prediction of externalizing behavior varied as a function of children’s and parents’ genders, as well as the attachment pattern (e.g., security, avoidance). Taken together, this dissertation shows that a continuous measure of attachment can help uncover the complexity of different attachment patterns, and in turn, provide a more nuanced understanding on how such patterns affect children’s social, emotional, and psychopathological development.