In order to verify whether the performance of patients with temporal lobe epilepsy (TLE) were different from those of the normal controls in various cognitive tasks, we conducted a neuropsychological study in the two groups, excluding the presence of other confounding factors in the group of TLE patients (newly referred, previously untreated, without cerebral lesions at neuroimaging, with low recurrence of seizures). In basal condition, no significant difference was found between the nine healthy controls and the eight TLE patients. The second aim of the study was to investigate the effect of the first administration of 400 mg controlled-release carbamazepine (CBZ-CR). The drug caused a significant decrease in verbal memory (short and long-term) in TLE patients, whereas, in the control group, the first administration of CBZ-CR caused only a less significant impairment of long-term verbal-memory. Finally, the cognitive effects of CBZ-CR were assessed after one month of therapy, only in the group of patients. Our data suggest that: 1) differences between patients and controls in basal conditions are not relevant, once the influence of seizures, of treatment and of anatomical lesions has been ruled out. In other words, our patients did not perform differently from controls simply because they had become epileptic 2) also in the controlled-release preparation, adverse effects of CBZ on cognitive function are minimal and limited to the beginning of treatment.