The most marked ' morning ' and 'evening ' types in a psychology class were identified by means of a questionnaire, and asked to record their oral temperatures and food intakes throughout the day during a 4-week and a 4-day period respectively. The morning group had its mean circadian temperature maximum 5 h earlier than the evening group, and had its cumulative food intake distribution curve 1 h ahead of the evening group. After adjusting the food distributions by 1 h in the time base to get a least-square fit, significant differences between the distributions remained. It is suggested that morning types have a more autonomous 24-hour-periodicity than evening types. It is concluded that the questionnaires have the power to discriminate extreme morning and evening types of individuals in terms of oral temperature and food intake. Food intake seems to be a sensitive enough measure to be included in studies of inter-individual differences of circadian rhythms.