Children identified as Attention Deficit Disorder (ADD) on the basis of teacher ratings using the SNAP Rating Scale were compared to non-ADD children on objective classroom measures including observations of classroom behavior, examination of the organization of children's desks, and scoring of daily academic work. Analyses provided some support for the ability of the objective measures to differentiate between teacher-identified ADD and non-ADD children, though there was considerable overlap on the distributions of most variables between groups. Six measures chosen by a discriminant analysis combined to predict teacher ratings in 83% of the cases. The inclusion of academic, observational, and desk measures in the discriminant function indicates the importance of a multivariate assessment of this construct.