We correlated the postresection electrocorticograms (ECoGs) of 80 patients who underwent temporal lobectomy under general anesthesia for treatment of intractable complex partial seizures with surgical results in three groups: seizure/aura free (32 patients), auras only (16 patients), and one or more postoperative seizures (32 patients) at mean follow-up times of 34,31, and 38 months, respectively. Spontaneous “residual spikes,” ie, present after all resections, were present in 47% of patients who had no postoperative seizures or auras. However, they occurred in 72% of patients with any postoperative seizures (p < 0.05). The location (convexity, mesial, or edge of resections) or the distribution (unifocal versus multifocal) of the residual spikes was not of prognostic value. Quantitative studies in 5-minute epochs of the postexcision ECoGs did not reveal a significant difference in the morphology of the residual spikes, ie, the amplitude or firing pattern (single versus polyspike), in the three groups. The group with postoperative seizures showed a higher number of spikes per epoch (≥50), but it was not significant. Although the study shows that patients with residual spikes may have good prognosis, they are at significantly higher risk for postoperative seizures as compared with those without residual spikes. The possibility that intensity of firing of residual spikes may be an additional predictor of outcome warrants further study.