Objective: To discover the outcomes for children placed late for adoption (between 5 and 11 years old) from public care and to establish which factors predict poorer outcome. Method: Data were collected prospectively (1993-2003) from a representative sample of domestic U.K. adoptive placements (N = 108) at the start of placement, at 1 year, and 6 years later. Most of the children entered care because of abuse and neglect. Outcome was assessed by the disruption rate, by a twofold classification of the character of continuing placements, and by an assessment of psychological well-being. Results: At the adolescent follow-up, 23% of placements had disrupted, 49% were continuing positively, and 28% were continuing but with substantial ongoing difficulties. Four factors contributed independently to a higher risk of disruption: older age at placement (odds ratio = 1.07), having been singled out from siblings and rejected (5.87), time in care (1.04), and a high level of behavioral problems (1.25). Two factors predicted differences in continuing placements. Conclusions: Late adoption can be successful in that half the children made good progress, but the extent of disruptions and difficulties in continuing placements gives rise to concern. Knowledge of predictors will help in devising planning pre- and postplacement support services.