Thirteen-month-old maltreated infants (n = 137) and their mothers were randomly assigned to one of three conditions: child-parent psychotherapy (CPP), psychoeducational parenting intervention (PPI), or community standard (CS). A fourth group of nonmaltreated infants (n = 52) and their mothers served as a nonmaltreated comparison (NC) group. A prior investigation found that the CPP and the PPI groups demonstrated substantial increases in secure attachment at postintervention, whereas this change was not found in the CS and the NC groups. The current investigation involved the analysis of data obtained at a follow-up assessment that occurred 12 months after the completion of treatment. At follow-up, children in the CPP group had higher rates of secure and lower rates of disorganized attachment than did children in the PPI or the CS group. Rates of disorganized attachment did not differ between the CPP and the NC groups. Intention to treat analyses also showed higher rates of secure attachment at follow-up in the CPP group relative to the PPI and the CS groups. However, groups did not differ on disorganized attachment. Both primary and intention to treat analyses demonstrated that maternal-reported child behavior problems did not differ among the four groups at the follow-up assessment. This is the first investigation to demonstrate sustained attachment security in maltreated children 12 months after the completion of an attachment theory informed intervention. The findings also suggest that, although effective in the short term, parenting interventions alone may not be effective in maintaining secure attachment in children over time.