The present study examined the diagnostic sensitivity of the WAIS relative to a modified Halstead-Reitan test battery by comparing the accuracy of these instruments in differentiating pseudoneurologic control subjects from two groups of epileptics which were believed to differ in degree of underlying brain pathology. A total of 96 subjects were equally represented in the control, mildly impaired, and moderately impaired study groups. Separate three-group and two-group discriminant function analyses were performed for the two sets of measures. Comparisons of kappa coefficients indicated that the WAIS was inferior in classification accuracy to the neuropsychological tests in making the three-group discriminations and in separating the mildly impaired subjects from the moderately impaired group. These findings suggest that the WAIS may be inferior in discriminatory power relative to a wider array of neuropsychological measures than has heretofore been evaluated. Furthermore, these data imply that the degree of underlying brain involvement associated with a seizure disorder may be determined best from neuropsychological tests rather than from a general measure of intellectual ability, such as the WAIS.