This paper presents a clinical and electrophysiological analysis of type and duration of seizures recorded by means of long-term video electroencephalogram (EEG) monitoring, a method which enables accurate diagnosis of psychogenic pseudoepileptic seizures occurring with or without epileptic seizures. Analysis is based on 1083 patients, hospitalized at our department between 1990 and 1997, with a preliminary diagnosis of epilepsy. Psychogenic pseudoepileptic seizures were diagnosed in 85 patients (7.8%). In 48 patients, pseudoepileptic seizures alone were diagnosed (group 1). whereas 37 patients had a mixed condition in which pseudoepileptic seizures were accompanied by epileptic seizures (group 2). For comparison of duration of pseudo- and epileptic seizures a control group (group 3), consisting of 55 patients randomly selected from the population of patients suffering from epileptic seizures alone, was parceled out. Long-term video EEG monitoring was performed in 70 patients. In 55 (79%) of these patients 230 seizures (221 pseudoepileptic and nine epileptic) were recorded. In 30 patients (32%), the diagnosis was based on clinical observation of the seizures and on the number of EEG recordings, including activating procedures such as sleep deprivation, photostimulation, hyperventilation and anti-epileptic drug withdrawal. We found that the duration of epileptic seizures was significantly shorter than the duration of psychogenic pseudoepileptic seizures. Our study has exposed the difficulties involved in the diagnosis of psychogenic pseudoepileptic seizures and the negligible value of neuroimaging techniques and interictal EEG recordings in the differential diagnosis of epileptic versus nonepileptic seizures. In this study, psychogenic seizures were significantly more frequent in women than in men; patient history analysis did not confirm the hypothesis that sexual abuse may cause psychogenic seizures.