We conducted a retrospective study to evaluate the efficacy of levetiracetam as adjunctive therapy in patients with refractory partial onset seizures, and in particular to assess effects following failed epilepsy surgery. Eighty-two patients with uncontrolled partial-onset seizures treated with levetiracetam were identified, of whom 21 failed prior epilepsy surgery (Group I) and 61 had no prior surgery (Group II). Group I patients (11 female, 10 male) ranged in age from 20 to 80 years (mean=40.7); mean age at onset of seizures was 14.4 years. The 61 patients in Group II (37 female, 24 male) ranged in age from 18 to 80 years (mean=41.5) and had mean age of onset of seizures 18.2 years. Patients who had ≥50 % reduction in seizure frequency were considered responders; the remaining patients were considered non-responders. In Group I, responder rate was 76.1 %: 10 patients (47.6 %) were seizures - free, and 6 patients (28.5 %) had ~ 50 % reduction in seizure frequency. In Group II, responder rate was 34.3 %: 9 patients (14.7 %) were seizure - free and 12 patients (19.6 %) had ≥ 50 % reduction. In Group I, 11 of 12 temporal resection patients (91.6 %) were responders, of whom 8 were seizure-free; of the remaining 9 operated patients, 5 (55.5 %) were responders and 2 patients were seizure-free. Three responders (14.2 %) in Group I developed a severe, delayed psychotic syndrome 4 9 months following levetiracetam introduction, leading to its discontinuation. This impressive data shows adjunctive levetiracetam therapy should be considered early following failed epilepsy surgery (especially temporal resection), and may have implications for its use before surgical intervention. Patients should be under close psychiatric observation in this clinical setting.