Objective and subjective cognitive measures were evaluated in 401 patients before and 3 and 6 months after introducing levetiracetam (LEV). Initially, cognitive impairment was indicated in 37-44% of the patients, and subjective impairments in 32-67%. With LEV, 87% of untreated patients changed to monotherapy, and 94% changed from mono- to polytherapy. The rate of retention of LEV was 97%; adverse events were reported by 7%. Under LEV, 36% achieved early seizure control, 25% achieved late seizure control, 33% continued to have seizures, and 7% had a relapse. Very good tolerance was reported by 68%, and cognitive improvement by 58%. Objective improvement was significant in 23-29% of the patients; 5-6% deteriorated. Better baseline scores, later-onset epilepsies, fewer initial antiepileptic drugs, and seizure control were predictive of a better cognitive outcome. Considering the uncontrolled study design and the increase in total drug load, LEV appeared safe and efficacious, and a general subjective and objective cognitive improvement could be noted. However, a controlled study design would be required to attribute these improvements to a specific drug action.