Time-use information is preferably obtained from diaries, as this method is considered more reliable than information from questionnaires. Data from the Danish Time Use Survey 2001 thus indicate differences in the level of unpaid work, whereas only minor differences appear for paid work. That is: people reporting many hours of paid work tend to over-report the actual number of hours worked, while those reporting a small number of hours tend to underreport their contribution. For unpaid work, the same pattern appears. Moreover, men are found to be more unreliable than women in evaluating their amount of work on the labour market, while the opposite is the case for the unpaid/household work, with women underreporting their contribution more than men. The implication is that labour supply studies based on questionnaire-information, i.e. Labour Force Surveys, are less accurate than studies based on diary-information.