“…If the two decision criteria, namely (1) dif ferential sensitivity of RT and RS, and (2) reliability are both taken into consideration, then the normally used 5-min recording duration is an optimal choice for PEEG recordings. If, however, the aim of the study is to measure the transition time from the wake state into sleep, then a different method should be applied such as, for example, the multiple sleep latency test [ 10,11], For the vigilance control, which was assured in our recording by a choice reaction paradigm, alternative tech niques have been proposed such as, for example (1) arous ing the subject if EEG signs of a vigilance decrease are seen by the supervising technician [ 12]; (2) buzzer control [ 13], and (3) verbal/cognitive tasks, such as constructing a story based on a few key words presented just before the recording session [14], The choice of the recording dura tion will then depend critically on the degree of involve ment demanded by the specific technique used to control the subjects' vigilance state.…”