Juvenile delinquency remains a major concern, with childhood adversity recognised as an important risk factor. While research links childhood adversity to delinquency, less is known about how different operationalisations of adversity compare or how multiple adversities co-occur in predicting delinquency. Many studies overlook key features of adversity, such as its type, accumulation, timing, and chronicity. This study addresses these gaps by comparing three approaches to operationalising childhood adversity: individual adversities, a cumulative adversity score, and latent class analysis, and examining their associations with delinquency. Data were drawn from the Millennium Cohort Study, a large, nationally representative longitudinal sample of young people in the UK (N=9,980). Fourteen types of childhood adversity were assessed, covering a broad range of threats to children's safety and support. Our findings show that, when operationalised as individual adversities, maternal substance misuse and exposure to school and street violence were the strongest predictors of delinquency. Latent class analysis identified nine distinct groups, including those characterised by low adolescent adversity, persistent poverty and low adolescent adversity, maternal substance misuse, and high overall adversity. These groups showed significant associations with delinquency. The cumulative adversity scores were also positively associated with delinquency. Overall, the findings highlight that young people exposed to adversity during adolescence or to chronic adversities are particularly vulnerable to engaging in delinquency. Importantly, each approach to operationalising adversity offered unique insights, demonstrating how our understanding of the relationship between childhood adversity and delinquency can differ depending on how adversity is conceptualised and assessed.