Background Mortality, including suicide and accidents, is elevated in self-harm populations. Although risk factors for suicide following self-harm are often investigated, rarely have those for accidents been studied. Our aim was to compare risk factors for suicide and accidents. Method A prospective cohort (n=30 202) from the Multicentre Study of Self-harm in England, 2000-2007, was followed up to 2010 using national death registers. Risk factors for suicide (intentional self-harm and undetermined intent) and accidents (narcotic poisoning, non-narcotic poisoning, and non-poisoning) following the last hospital presentation for self-harm were estimated using Cox models. Results During follow-up, 1833 individuals died, 378 (20.6%) by suicide and 242 (13.2%) by accidents. Independent predictors of both suicide and accidents were: male gender, age ≥35 years (except accidental narcotic poisoning) and psychiatric treatment (except accidental narcotic poisoning). Factors differentiating suicide from accident risk were previous self-harm, last Method of self-harm (twofold increased risks for cutting and violent self-injury versus self-poisoning) and mental health problems. A risk factor specific to accidental narcotic poisoning was recreational/illicit drug problems, and a risk factor specific to accidental non-narcotic poisoning and non-poisoning accidents was alcohol involvement with self-harm. Conclusions The similarity of risk factors for suicide and accidents indicates common experiences of socio-economic disadvantage, life problems and psychopathology resulting in a variety of self-destructive behaviour. Of factors associated with the accidental death groups, those for non-narcotic poisoning and other accidents were most similar to suicide; differences seemed to be related to criteria coroners use in reaching verdicts. Our findings support the idea of a continuum of premature death. © 2011 Cambridge University Press.