A widely used statistical method to test for genetic association is the transmission disequilibrium test (TDT) using two parent-proband trios. West et al 1 have presented evidence from clinically ascertained ADHD families that children from trios were less likely to have DSM-IV combined subtype ADHD and conduct disorder. They suggest that the exclusion of parentproband duos could reduce the power of the TDT and similar tests to detect susceptibility genes for this subtype of ADHD. We sought to test this hypothesis in a population-based sample of twin families, while controlling for the effects of other proband and family characteristics in a multivariant logistic regression framework using both latent class and DSM-IV ADHD subtype definitions. For both latent class and DSM-IV defined combined and inattentive ADHD, sex of the proband and comorbid conduct disorder or oppositional defiant disorder, significantly predicted diagnosis. For latent class and DSM-IV defined combined subtype, younger age also significantly predicted ADHD subtype. Latent class and DSM-IV defined combined subtype ADHD with comorbid conduct disorder was significantly less common in children from trios while conduct disorder without ADHD did not differ in frequency between families with zero, one or two participating parents.