A principal components analysis was carried out on subjective data obtained from the sleep questionnaire systematically used in our laboratory on the morning following a polysomnographic recording. This technique, based on 70 healthy subjects, revealed three principal axes which explore the questionnaire. The first axis reflects the subjective evaluation of the quality of sleep. The second underlines the importance of dreams and of the brief waking periods which immediately follow. Finally, the third axis brings out the subjective assessment of sleep efficiency. The third axis was the most ambiguous and the only one to show a gender difference. Women tend to better estimate general sleep events, while men tend to evaluate better the more circumstantial events. Finally, given the results, it would be interesting to apply a similar analysis to populations presenting sleep disorders in order to establish the presence of an eventual pattern specific to the disorder.