OBJECTIVE: We aimed to compare long-term social outcomes in young adults with childhood-onset epilepsy (cases) with neurologically normal sibling controls. METHODS: Long-term social outcomes were assessed at the 15-year follow-up of the Connecticut Study of Epilepsy, a community-based prospective cohort study of children with newly diagnosed epilepsy. Young adults with childhood-onset epilepsy with complicated (abnormal neurologic exam findings, abnormal brain imaging with lesion referable to epilepsy, intellectual disability (ID; IQ < 60) or informative history of neurologic insults to which the occurrence of epilepsy might be attributed), and uncomplicated epilepsy presentations were compared to healthy sibling controls. Age, gender, and matched-pair adjusted generalized linear models stratified by complicated epilepsy and 5-year seizure-free status estimated adjusted odds ratios (aORs) and 95% confidence intervals [CIs] for each outcome. RESULTS: The 15-year follow-up included 361 individuals with epilepsy (59% of initial cases; N = 291 uncomplicated and N = 70 complicated epilepsy; mean age 22 years [standard deviation, SD 3.5]; mean epilepsy onset 6.2 years [SD 3.9]) and 173 controls. Social outcomes for cases with uncomplicated epilepsy with >/=5 years terminal remission were comparable to controls; cases with uncomplicated epilepsy <5 years seizure-free were more likely to be less productive (school/employment < 20 h/week) (aOR 3.63, 95% CI 1.83-7.20) and not to have a driver's license (aOR 6.25, 95% CI 2.85-13.72). Complicated cases with epilepsy <5 years seizure-free had worse outcomes across multiple domains; including not graduating high school (aOR 24.97, 95% CI 7.49-83.30), being un- or underemployed (<20 h/week) (aOR 11.06, 95% CI 4.44-27.57), being less productively engaged (aOR 15.71, 95% CI 6.88-35.88), and not living independently (aOR 10.24, 95% CI 3.98-26.36). Complicated cases without ID (N = 36) had worse outcomes with respect to productive engagement (aOR 6.02; 95% CI 2.48-14.58) compared to controls. Cases with complicated epilepsy were less likely to be driving compared to controls, irrespective of remission status or ID. SIGNIFICANCE: In individuals with uncomplicated childhood-onset epilepsy presentations and 5-year terminal remission, young adult social outcomes are comparable to those of sibling controls. Complicated epilepsy, notable for intellectual disability, and seizure remission status are important prognostic indicators for long-term young adult social outcomes in childhood-onset epilepsy.