Because Wada evaluations are not standardized, it is impossible to know to what degree method variance accounts for reported differences in results. To examine this problem, three comprehensive epilepsy surgery centers compared the efficacy of two Wada memory methods to predict seizure onset laterality in 152 children being considered for epilepsy surgery. Wada memory asymmetries were evaluated using either real objects with no verbal response required or more mixed stimuli requiring a verbal response. When using real objects, Wada memory performance was significantly worse when relying on the side of seizure onset in both left and right seizure onset children. In contrast, Wada memory performance using mixed stimuli was worse on the side of seizure onset only among patients with seizures originating in the left-hemisphere. The superiority of real objects was most apparent in younger children with left side seizure onset. Results suggest the use of mixed stimuli is less sensitive to the effects of unilateral seizure onset, and thus, diminishes the capacity of the Wada test to predict lateralized seizure onset in children.