With increasing awareness that ADHD is chronically disabling, a burgeoning literature has examined childhood clinical indicators of ADHD persistence. This study investigates whether childhood factors reflecting biological risk and cognitive reserve have additive predictive value for the persistence of ADHD that is unique beyond childhood indicators of disorder severity. One-hundred thirty children with ADHD (mean age = 8.9 years, 75 % male) were followed into adolescence (mean age = 14.0 years). Childhood ADHD and co-morbidities were assessed via interviews with parents and teachers; parental psychopathology was assessed via parent interview; exposure to neurobiological and psychosocial adversity were indexed by parent questionnaire; and cognitive reserve was evaluated through children’s performance on measures of IQ and executive functioning. Univariate analyses identified childhood inattention and hyperactivity-impulsivity, co-morbid oppositional defiant disorder, overall impairment, and paternal anxiety and depression as more prevalent amongst adolescents with persistent compared with remitted ADHD. Only child-level predictors remained significant in a final multivariate model. These results suggest that children who are most likely to experience persistent ADHD have a more severe clinical presentation in childhood, reflected by increased levels of inattention, oppositional behavior, and impairment. They also are more likely to have fathers with internalizing concerns, but these concerns do not uniquely predict ADHD persistence beyond child-level factors. Contrary to expectations, childhood adversity and cognitive functioning did not predict the course of ADHD.