Using data from the Montreal Longitudinal-Experimental Study, we examined the impact of poverty (and its correlate, family configuration status) on academic placement and self-reported delinquency in boys at age 16. We then investigated whether the relation between family economic hardship and antisocial behaviour is direct or indirect by considering the value of parenting practices and academic failure as process variables in the model. Data included official records, and parent, teacher, and self-reports. The temporal intensity of poverty was classified into five categories: never-poor; always-poor; poor-earlier; poor-later; and transitory-poverty. Family configuration status was classified by both temporal characteristics and number of marital transitions: intact-family; short-term-single; long-term-single; short-term-remarried; long-term-remarried; and multiple-marital-transitions. Results revealed that when maternal education and early childhood behaviour were controlled, poverty had an effect on both academic failure and extreme delinquency. This effect was independent of family configuration status. Although they both significantly predicted extreme delinquency on their own, academic failure and parental supervision did not mediate the relationship between poverty and delinquency. Divorce increased the risk of theft and fighting at age 16, regardless of financial hardship. Parental supervision only helped explain the effects of divorce on boys' fighting.