Part 1 provides confirmatory evidence that relatively independent AAI dismissing and preoccupied states of mind—along with variation in inferred maternal and paternal experience—capture the full range of participants’ AAI discourse. Taxometric analyses demonstrated that individual differences are more accurately represented dimensionally than categorically. Part 2 reports evidence of weak but statistically significant stability in attachment from infancy through late adolescence, and lawful sources of continuity and change over time—maternal sensitivity, father absence, paternal depression, and negative life events. A specific focus on individuals who described below-average childhood experiences in the AAI but did so in a coherent manner (i.e., "earned-secures") replicated evidence that they actually received average or better parental care, but also experienced significant family stressors in childhood. Additional analyses suggested theory-consistent developmental antecedents of the four AAI dimensions (i.e., dismissing, preoccupied, inferred maternal and paternal experiences). Together, these results represent a significant step forward in our understanding of adult attachment and its origins.