Neuropsychological assessment has been extensively used in the presurgical evaluation of temporal lobe epilepsy (TLE) patients to assist in determining lateralization of seizure onset. Very few studies have examined the accuracy with which commonly used neuropsychological instruments provide this information in the individual patient. In 81 patients (49 right-, 32 left-TLE) without space-occupying lesions in whom correct seizure lateralization was inferred on the basis of postsurgical seizure-free status, we compared the frequency with which discrepancies between the Wechsler Memory Scale–Revised (WMS–R; Wechsler, 1987) Verbal and Visual Memory Indices, Warrington Recognition Memory Test (WRMT; Warrington, 1984) Words and Faces scaled scores, and Intracarotid Amobarbital Procedure (IAP) hemispheric memory scores correctly predicted seizure lateralization in the individual patient. Using both clinical analysis and discriminant function analyses (DFA), the IAP was found to be a clearly superior predictor of seizure laterality to the neuropsychological measures, whether used individually or in combination with one another. Using clinical analysis the WRMT was found to be a superior predictor to the WMS–R, which frequently gave false lateralizing information. Using all 3 measures in combination with one another, 87.1% of patients were correctly lateralized using DFA. Correctly lateralized patients were older and had longer durations of seizure disorder. (JINS, 1997, 3, 159–168.)