Realism about mental disorders is a perennial area of dispute, but the controversy burns especially intensely for Attention Deficit Hyperactivity Disorder (ADHD). In this dissertation, I clarify what is at issue in these debates, surveying how realists have typically argued for mental disorder realism: the definitional debate about health and illness. I argue that the realist need not be committed to the terms of the definitional debate and recommend that a better approach is to show that mental disorders are natural kinds. While there are many accounts of kind-hood on offer, I adopt Richard Boyd’s homeostatic property cluster (HPC) theory of kinds, which I interpret through the philosophy of neuroscience literature on mechanisms. In sum, I conclude that if ADHD is a natural kind – and thus real – then individuals diagnosed with the disorder should be sufficiently similar with respect to an underlying cognitive neurobiological mechanism. To determine whether ADHD individuals are similar in this way, I consider the question through Russell Barkley’s Executive Function Model of ADHD. Relying primarily on the cognitive neurobiological research, I argue that there is now reasonable evidence to conclude that the DSM classification of ADHD corresponds not to a single natural kind, but several. Thus, ADHD is thus real.