To examine Swanson, Nolan, and Pelham-IV (SNAP-IV) psychometric properties, parent (N = 1,613) and teacher (N = 1,205) data were collected from a random elementary school student sample in a longitudinal attention deficit hyperactivity disorder (ADHD) detection study. SNAP-IV reliability was acceptable. Factor structure indicated two ADHD factors and an oppositional defiant disorder (ODD) factor. Parent and teacher scores varied by gender and poverty status (d = .49-.56) but not age; only teacher scores varied by race (d = .25-.55). Screening and diagnostic utility was evaluated with likelihood ratios (LRs) and posttest probabilities. Parent SNAP-IV scores above 1.2 increased probability of concern (LR > 10) and above 1.8, of ADHD diagnosis (LR > 3). Teacher hyperactivity/impulsivity scores above 1.2 and inattention scores above 1.8 increased probabilities of concern only (LR = 4.2 and >5, respectively). Higher teacher scores for African American children and race differences in measurement models require future study.