Background: No ecological studies have examined the relationship between area characteristics, individual characteristics and self-harm repetition. Aims: To investigate the association between area-level factors and incidence and repetition of self-harm, and to identify which area-level factors are independently associated with repetition after adjustment for individual factors. Method: Prospective cohort study using the Manchester Self-Harm database. Adults who were resident in Manchester and presented to an emergency department following self-harm between 1997 and 2002 were included (n=4743). The main outcome measure was repeat self-harm within 6 months of the index episode. Results: Four individual factors (previous self-harm, previous psychiatric treatment, employment status, marital status) and one area-based factor (proportion of individuals who were of White ethnicity) were independently associated with repetition. Conclusions: Repetition of self-harm may be more strongly related to individual factors than to area characteristics. We need to better understand the processes underlying ecological associations with suicidal behaviour before embarking on area-based interventions.