Attention-deficit/hyperactivity disorder (ADHD) is a clinically heterogeneous, highly heritable and genetically complex disorder. The pathways from genes to behaviour are still unknown. Endophenotypes or intermediate phenotypes that are more closely linked to the neurobiological substrate than the core symptoms of ADHD may help to disentangle these complex relationships between genes and behaviour and to clarify its etiology and pathophysiology. Heritability and stability (state independence) represent key components of any useful endophenotype. Various other criteria for the selection of useful endophenotypes have been proposed. A review of the current state of the research on potential endophenotypes for ADHD will be given.