--- - |2+ Background Only one‐third of young people who experience suicidal ideation attempt suicide. It is important to identify factors which differentiate those who attempt suicide from those who experience suicidal ideation but do not act on these thoughts. Methods Participants were 4,772 members of the Avon Longitudinal Study of Parents and Children (ALSPAC), a UK population‐based birth cohort. Suicide ideation and attempts were assessed at age 16 years via self‐report questionnaire. Multinomial regression was used to examine associations between factors that differentiated adolescents in three groups: no suicidal ideation or attempts, suicidal ideation only and suicide attempts. Analyses were conducted on an imputed data set based on those with complete outcome data (suicidal thoughts and attempts) at age 16 years (N = 4,772). Results The lifetime prevalence of suicidal ideation and attempts in the sample was 9.6% and 6.8% respectively. Compared to adolescents who had experienced suicidal ideation, those who attempted suicide were more likely to report exposure to self‐harm in others (adjusted OR for family member self‐harm: 1.95, for friend self‐harm: 2.61 and for both family and friend self‐harm: 5.26). They were also more likely to have a psychiatric disorder (adjusted OR for depression: 3.63; adjusted OR for anxiety disorder: 2.20; adjusted OR for behavioural disorder: 2.90). Other risk factors included female gender, lower IQ, higher impulsivity, higher intensity seeking, lower conscientiousness, a greater number of life events, body dissatisfaction, hopelessness, smoking and illicit drug use (excluding cannabis). Conclusions The extent of exposure to self‐harm in others and the presence of psychiatric disorder most clearly differentiate adolescents who attempt suicide from those who only experience suicidal ideation. Further longitudinal research is needed to explore whether these risk factors predict progression from suicidal ideation to attempts over time. - Journal of Child Psychology and Psychiatry, EarlyView.