The clinical, electroencephalographic and Imaging diagnostic features of the 45 patients of Localization-related Epilepsy who had elementary visual symptoms at seizure onset were investigated. There were 24 males and 21 females aged 19 to 78. Their ages at seizure onset ranged from 1 to 55 with a mean of 15.4. The patients were divided into 3 groups based on the pattern of elementary visual symptoms: Group 1; 23 patients (51%) with only elementary positive visual symptoms, Group 2; 9 patients (20%) with only elementary negative visual symptoms, Group 3; 13 patients (29%) with others. It came to our notice that 8 (18%) patients in Group 3 had positive + negative (P*N) visual seizures such as scintillation scotoma. It was quite difficult to draw a distinction between the P*N seizures and scintillation scotoma in migrainous patients in quality. Therefore, a question arises whether similarities between them are ascribed to the same underlying mechanism. It is possible that the mechanism of the P*N seizures is different from that of the scintillation scotoma, but the both produce the same condition. However, we are not competent to discuss this hypothesis. Further work along this line is necessary. In addition to the elementary visual symptoms, autonomic (69%), focal motor (29%), illusion (29%), vertiginous (22%) manifestations, etc. were also observed. These manifestations suggest that epileptic ictal discharges spread into many different brain areas. Regarding Group 1-3, occurrence of illusional seizures was more common in Group 2. In this group, in only one patient, occipital interictal discharges were observed. The subjects were subclassified into two groups depending on whether the seizures were well controlled (good outcome) or poorly controlled (poor outcome group). On these groups, comparative studies were performed. The poportion of the patients with a family history of convulsive disorder was higher in the latter than that in the former. To the contrary, the poportion of the patients with a etiologic episode was higher in the former than that in the latter. With respect to the imaging study, regardless of localizations, abnormal findings were detected in 25% on CT scan, in 29% on MRI, and in 71% on SPECT. Whether or not these abnormal findings completely or partially agreed with presumed epileptogenic brain region (occipital lobes) was investigated. CT scan was positive (correspondent) in 8%, MRI was positive in 13%, and SPECT was positive in 54%. The incidence of SPECT abnormalities was higher in the poor outcome group (70 %) than that in the good outcome one (43%). We may, therefore, conclude that SPECT seems to be useful for the detection of epileptogenic region, especially in intractable cases. But the incidence of these imaging studies' abnormalities located in occipital lobe was not so high, therefore, there seems to be no doubt that clinical symptoms and EEG findings are most important for diagnosis.