Between January 1992 and July 1998, 320 patients were presurgically evaluated for medically refractory epilepsy at the University Hospital of Gent. All patients underwent a comprehensive presurgical evaluation, including extensive neurological history and examination, video-EEG monitoring of interictal EEG and habitual seizures, and optimum magnetic resonance (MR). In a large subgroup of these patients, a comprehensive neuropsychological examination and interictal 18FDG-PET were performed. Subsequently, a bilateral carotid angiography and intracarotid amytal procedure (Wada-test) were planned in 49 patients to establish hemispheric language dominance and bilateral memory function. After proper selection, 23 patients underwent invasive video-EEG monitoring with intracranial implantation of parenchymal and/or subdural electrodes to further document the area of seizure onset. From the initial group of 320 potential surgical candidates, 75 patients (42 males, 33 females) with mean age of 29 years (range: 2 months-55 years) and mean duration of uncontrolled seizures of 15 years (range: 2 weeks-38 years) eventually underwent a surgical procedure. Sixty of 75 patients were on high dose antiepileptic polytherapy. Optimum MR detected structural abnormalities, confined to a limited brain area, in 71 patients. These abnormalities were of space-occupying nature in 31 cases; an atrophic lesion was suspected in 39 patients; a combination of space-occupying and atrophic lesion was seen in 1 case. Structural abnormalities were most frequently located in the temporal lobe (n = 53) and the frontal lobe (n = 10). Video-EEG monitoring documented complex partial seizures in 67 patients with occasional secondary generalisation in 32. Most patients had complex partial seizures of temporal lobe as defined by clinical and EEG criteria. Two patients had only simple partial seizures. Ultimately, an area of seizure onset could be determined in all patients. Temporal lobectomy with hippocampectomy was the most commonly performed procedure (n = 42). In 13 patients, complete lesionectomies were performed for epileptogenic structural lesions in and outside the temporal lobe. In 2 patients, only partial lesionectomies were possible; in 5 patients, only biopsies in combination with partial lesionectomies could be performed. Anterior 2/3 callosotomy was performed in 4 patients and hemispherectomy was performed in 2 patients. Postsurgical seizure control, after average follow-up of 50 months (range: 12-98 months), was excellent in 49 patients who became seizure-free. In these patients, antiepileptic therapy was tapered 2 years after surgery. Patients in whom only biopsies or partial lesionectomies were performed have poor seizure control. Epilepsy surgery is a rewarding therapeutic alternative for patients with medically refractory epilepsy. Comprehensive presurgical evaluation and epilepsy surgery provide excellent neurological, neurophysiological, neuropsychological and imaging research opportunities.