Attention-Deficit/Hyperactivity Disorder (ADHD) diagnosed in childhood is associated with adverse socioeconomic outcomes, but questions remain about causality. Prior work is limited by single-timepoint assessments, narrow socioeconomic measurement, and inadequate control for genetic and environmental confounds. We used data from the Minnesota Twin Family Study, a multi-decade longitudinal study of 2764 twins, to examine associations between ADHD and socioeconomic status (SES) across development. Childhood ADHD was defined as meeting diagnostic criteria at baseline (n = 131). SES was measured with parental education and occupation, household income, and neighborhood disadvantage in childhood, and participants' own education, occupation, and income in adulthood. Co-twin control analyses accounted for genetic and shared environmental confounding. We examined apparent adult-onset ADHD (n = 32) as an exploratory comparison. Childhood ADHD was associated with both rearing family SES and adult education. In twin pairs discordant for childhood ADHD, the affected twin showed lower levels of education, indicating that the association does not owe entirely to the intergenerational transmission of SES and consistent with a causal effect of ADHD. Within-pair differences in adult occupational status and income were not significant, suggesting these associations reflect shared familial factors. Adult-onset ADHD showed weaker associations overall, with no significant within-pair effects. Lower SES in childhood was associated with childhood ADHD, which predicted lower adult SES, particularly for education. Co-twin analyses suggest the childhood ADHD-educational attainment association is not entirely attributable to genetic or familial confounds, whereas adult-onset ADHD showed weaker and less consistent patterns.