The aim of the present study was to find out if Adam's (1994) developmental model of suicidal behaviours was applicable to self-harming behaviours in adolescence. It was hypothesised that, after controlling for level of depression, insecure attachment would predict frequency of self-harm and that this effect would be mediated by dissatisfaction with perceived social support. The subjects (n = 20) were drawn from new referrals to a mental health service for adolescents in Edinburgh. A number of questionnaires were administered to the young people within the format of a structured interview. The Relationship Questionnaire and the Relationship Scales Questionnaire were used to measure attachment, the Significant Others Scale measured satisfaction with perceived social support and the Beck Depression Inventory was used to measure level of depression. Several multiple regression analyses were used to build up an observed variable path model. It was found that there was no significant relationship between attachment and self-harm. However, both dissatisfaction with social support and level of depression were found to significantly predict self-harm. Fearful attachment style was found to significantly predict both satisfaction with social support and level of depression. The underlying dimensions of attachment styles were also considered. Model of self was found to significantly predict level of depression and model of other was found to predict satisfaction with social support. The results of the present study do no suggest that Adam's ( 1994) model of suicidal behaviours is applicable to self-harming behaviours in adolescence. It is suggested that a developmental model may be more applicable to those individuals who continue self-harming in to adulthood. The limitations of the study are discussed and therefore how research in this area might usefully progress.