Non-epileptic seizures (NES) are reported in 18-23% of patients referred to comprehensive epilepsy centres. Non-epileptic seizures may also be present in 5-20% of the patients who are diagonised as having refractory seizures. Because of their prevalence, financial and psychosocial outcomes cannot be ignored and accurate diagnosis is of the utmost importance. Various methods of seizure induction have been developed with the aim of differentiating epileptic from non-epileptic seizures. However, recording the attacks by video-EEG monitoring is the gold standard. In our outpatient EEG laboratory we try to induce seizures with verbal suggestion or IV saline infusion in patients who are referred by a clinician with the diagnosis of probable nonepileptic seizures. In this study we investigated the results of 72 patients who were referred between January 1992-June 1996. Non-epileptic seizures were observed in 52 (72.2%) patients. Thirteen of these patients still had risk factors for epilepsy. We could not decide whether all of their previous attacks were non-epileptic because 10-30% of the patients with NES also have epileptic seizures. For a more accurate diagnosis it was decided that these 13 patients, together with the 20 patients who did not have seizures with induction, needed video-EEG monitoring. Thirty-nine patients who had NES and no risk factors for epilepsy were thought to have pure non-epileptic seizures. We claim that not all patients suspected of having NES need long-term video-BEG monitoring and almost half (54.2%) of the cases can be eliminated by seizure induction with some provocative techniques.