The majority of the research assessing the differences in driving performance between medicated and non-medicated Attention Deficit Hyperactivity Disorder (ADHD) teenagers has relied on driving simulators and surveys. This pilot study was conducted in a real-world environment using a naturalistic driving methodology where continuous driving data were collected for 15 to 24 months (9 months of learner’s permit driving and at least 6 months of early licensure driving) using the Virginia Tech Transportation Institute’s NextGen data acquisition system. Ten teenaged drivers (15.5 and 16 years of age) with learner’s permits and diagnoses of ADHD were recruited, and their driving data were compared to a control group of 45 newly licensed teen drivers from the Supervised Practice Driving Study who scored 0 on the Disruptive Behavior Disorder Rating Scale rating scale, and thus were confirmed to be free of ADHD symptoms. The mean crash and near-crash (CNC) rates per 1,000 hours for the ADHD group and the control group were 22.59 and 11.53, respectively. The data were further analyzed to investigate driver behaviors that contributed to the occurrence of CNC events. Although both groups engaged in risky driver behaviors during CNC events, when the two groups were compared, the results showed that ADHD novice teen drivers engaged in risky driving behaviors, particularly those related to vehicle mishandling, more frequently than non-ADHD teens. The ADHD teens were also more distracted by engagement with high-risk secondary tasks, such as texting and passenger interaction, which have been found to significantly increase the risk of CNC involvement, than non-ADHD teen drivers.