Long-term electrocorticographic (ECoG) monitoring data from subdural strip electrodes are analyzed to determine factors associated with seizure-free outcome from anterior temporal lobectomy. A total of 89 consecutive patients with complex partial seizures, in whom long-term ictal video/scalp electroencephalographic monitoring was insufficient to localize their epileptogenic focus, were subsequently evaluated with long-term ictal ECoG monitoring using subdural strip electrodes. Each patient underwent anterior temporal lobectomy based on the ictal ECoG data and has been followed for at least 1 year. The following parameters were found to be statistically significant in predicting a seizure-free outcome: unilateral onset, electrical onset pattern beginning as fast spike trains, absence of frontal lobe background desynchronization at onset, and an interhemispheric propagation time of greater than 8 seconds. Electrocorticographic criteria that were not associated with seizure outcome included: right- versus left-sided onset, time from electrical to clinical ictal onset, focality of onset (number of strip electrode contacts involved), and stereotypical ECoG onset. When present, the interictal focus was concordant with the ictal focus in most patients (96%), but was falsely lateralizing in 4% of cases. It is suggested that these data should improve patient selection for temporal lobectomy when subdural strip monitoring is used during preoperative evaluation.