Our purpose is to determine predictors of outcome in patients with refractory temporal lobe epilepsy and normal high resolution magnetic resonance imaging (MRI) who undergo surgical therapy. We identified 23 patients who underwent temporal lobectomy and had normal pre-operative MRI, including surface coil phased array temporal lobe imaging. All were followed at least 2 years after surgery. We graded outcome as seizure-free, > 75% reduction in seizures, or < 75% reduction in seizures. We examined pre-operative interictal and ictal electroencephalographic (EEG) findings, age of onset, gender, duration of epilepsy, risk factors, family history, physical findings, age at operation, side of operation, and pathology of resected tissue in order to determine if any of these factors were associated with outcome. Overall, 48% (11/23) of patients were seizure-free, 39% (9/23) had > 75% reduction in seizures, while 13% (3/23) had < 75% reduction in seizures. Only the EEG findings were useful in predicting outcome. When ictal onsets arose from basal-temporal regions, 61% (11/18) of patients were seizure-free, while none (0/5) were seizure-free when seizures arose from mid-posterior temporal regions (P = 0.04). Interictally, if all epileptiform patterns were localized exclusively to one basal-temporal region, a finding that invariably correlated with ictal onsets, 78% (7/9) of patients were seizure-free, while only 29% (4/14) were seizure-free if discharges were bilateral or multifocal (P = 0.04). We conclude that surgery may be a reasonable treatment for some patients with intractable temporal lobe seizures and normal MRI. The best outcomes occur when seizure onsets and interictal epileptiform patterns are exclusive to one basal-temporal region. Unfavorable outcomes are most likely to occur when ictal origins are from mid-posterior temporal regions and when interictal discharges are bitemporal or multifocal in distribution.