We examined the prototype v. revisionist models of attachment stability with a five-wave, 6-year, longitudinal study of attachment security from middle childhood to adolescence in a White Western European sample (N = 157; Wave 1 M[subscript age] = 10.91, SD = 0.87; 52% female). Attachment was assessed using both questionnaire (Experiences in Close Relationships) and narrative-based measures (Attachment Script Assessment). In addition, a set of potential moderators of prototype-like stability were examined. Results indicated that data from both attachment assessments best fit the prototype model. Moderator analyses indicated that male sex significantly undermined the influence of an attachment prototype and parent-child conflict and parental divorce enhanced the influence of an attachment prototype on stability.