Cortical resection surgery for epilepsy is costly, involves risk, and does not result in significant seizure relief in a number of patients. Therefore, it is important to develop methods which will as accurately as possible identify in advance those persons who will or will not be helped by this procedure. From a review of the literature and from a presentation of original data, it is concluded that psychological/neuropsychological variables are valid predictors of cessation of seizures following surgery but that they are only infrequently used. A multidisciplinary, multivariate statistical approach, using EEG findings and psychological/neuropsychological variables in combination, provides the most accurate predictions which are superior to those derived from using either EEG variables or psychological/neuropsychological variables alone. This approach also avoids much of the predictive instability encountered when single variables are evaluated. Since psychological/neuropsychological variables are obtained at low cost and are non-invasive, further evaluation of their role as predictors of seizure relief following surgery is needed.