BACKGROUND Children in foster care constitute a risk population for developing symptoms of attachment disorders. However, little is known about the longitudinal course of attachment disorders and their association with attachment security in foster children. METHODS This longitudinal study assessed attachment disorder symptoms in a sample of foster children (n = 55) aged 12 to 82 months. Foster parents with a newly placed foster child were assessed at three points during the first year of placement. At all assessment points, the Disturbance of Attachment Interview (DAI; Smyke and Zeanah in Disturbances of attachment interview, Tulane University, New Orleans, 1999) and the Attachment Q-sort (AQS; Waters and Deane in Monogr Soc Res Child Dev 50:41-65, 1985 German version as reported (Schölmerich and Leyendecker in Deutsche Übersetzung des attachment behavior Q-Set, revision 3.2. Unpublished manual, Ruhr University Bochum, Bochum, 1999) were used to investigate the interplay between disorder symptoms and attachment security. RESULTS The results revealed that the symptoms of attachment disorders decreased. The decrease was more pronounced for the inhibited than for the disinhibited symptoms with marked changes in the first 6 months of placement. There was a noticeable gender difference in the development with boys showing a more pronounced decrease in inhibited attachment disorder symptoms and a stronger increase of attachment security. After 12 months, no significant gender effects were found. Regarding the association between symptoms of attachment disorders and attachment security, a significant negative correlation between the inhibited attachment disorder symptoms and attachment security was found 12 months after placement. CONCLUSIONS Attachment disorder symptoms decreased in the stable foster care environment. Thus, foster care seems to be an effective placement option regarding children's attachment development.