Young people who self-harm are at an increased risk of suicide. Furthering our understanding of the risk factors for self-harm is essential for identifying high-risk groups, which can be used to inform the design of preventative interventions. This study used the Avon Longitudinal Study of Parents and Children (ALSPAC) and applied latent class analysis to the risk factors for self-harm at ages 13 and 17. Longitudinal associations between the latent classes and self-harm at ages 17 and 20 were examined. Cross-cohort comparisons were conducted between this study and a previous study using Irish data. At age 13 there was a low risk group, a peer problems group, and substance use group, similar for the two cohort studies, and a family conflict group, which was the least similar group to its matching group in the Irish study. All of these age 13 high-risk groups had approximately twice the relative risk (between 1.3 and 2.5) for self-harm at age 17 compared to the low risk group. The age 17 models were very similar across the two cohorts, each with a low risk group, a depression and high substance use group, a depression and low substance use group, and a substance use group. The relative risk of self-harm at age 20 for these high-risk groups compared the low risk group ranged from 3.6 to 8.0. These groups could help identify those at risk of self-harm and inform the design of prevention programmes to reduce self-harm behaviour in young people.