Background: Self-harm (intentional self-poisoning and self-injury) in children and adolescents is often repeated and is associated with increased risk of future suicide. We have investigated factors associated with these outcomes. Method: We used data collected in the Multicentre Study of Self-harm in England on all self-harm hospital presentations by individuals aged 10-18 years between 2000 and 2007, and national death information on these individuals to the end of 2010. Cox hazard proportional models were used to identify independent and multivariable predictors of repetition of self-harm and of suicide. Results: Repetition of self-harm occurred in 27.3% of individuals (N = 3920) who presented between 2000 and 2005 and were followed up until 2007. Multivariate analysis showed that repetition was associated with age, self-cutting, and previous self-harm and psychiatric treatment. Of 51 deaths in individuals who presented between 2000 and 2007 and were followed up to 2010 (N = 5133) half (49.0%) were suicides. The method used was usually different to that used for self-harm. Multivariate analysis showed that suicide was associated with male gender [Hazard ratio (HR) = 2.4, 95% CI 1.2-4.8], self-cutting (HR = 2.1, 95% CI 1.1-3.7) and prior psychiatric treatment at initial presentation (HR = 4.2, 95% CI 1.7-10.5). It was also associated with self-cutting and history of psychiatric treatment at the last episode before death, and history of previous self harm. Conclusions: Self-cutting as a method of self-harm in children and adolescents conveys greater risk of suicide (and repetition) than self-poisoning although different methods are usually used for suicide. The findings underline the need for psychosocial assessment in all cases. © 2012 The Authors. Journal of Child Psychology and Psychiatry © 2012 Association for Child and Adolescent Mental Health.