Converging evidence from family, twin, and adoption studies points to a substantial genetic component of the etiology of attention deficit hyperactivity disorder (ADHD). These data about ADHD have motivated molecular genetic studies of the disorder, which have produced intriguing but somewhat conflicting results. Some studies have reported associations with candidate genes and others not. Our review of the literature shows that one problem facing molecular genetic studies of ADHD is that its recurrence risk to first-degree relatives is only about five times higher than the population prevalence. This suggests that, to produce consistently replicated results, molecular genetic studies should either use much larger samples or should select those families in which genes exert the largest effect. Risch [(1990a) Am J Hum Genet 46:222-228; (1990b) Am J Hum Genet 46:229-241] proved that the statistical power of a linkage study increases with the magnitude of risk ratios (Ξ»'s) computed by dividing the affection rate among each relative type to the rate of affection in the population. Our prior work suggests two dimensions of genetic heterogeneity that might be useful for selecting ADHD subjects for molecular genetic studies: comorbidity with con-