The European Community Directive on Working Time, which should have been implemented in member states of the European Community by November 1996, contains several requirements related to working hours, including the right of employees to refuse to work more than 48 hours a week. The United Kingdom government attempted to oppose the Directive, arguing that there is no convincing evidence that hours of work should be limited on health and safety grounds. Much of the research in this area has focused on the problems of shiftworking and previous reviews have therefore tended to emphasise this aspect of working hours. However, there is much less information about the effects of overtime work, which is a central element of the terms of the Directive. This paper reviews the current evidence relating to the potential effects on health and performance of extensions to the normal working day. Several gaps in the literature are identified. Research to date has been restricted to a limited range of health outcomes--namely, mental health and cardiovascular disorders. Other potential effects which are normally associated with stress--for example, gastrointestinal disorders, musculoskeletal disorders, and problems associated with depression of the immune system, have received little attention. Also, there have been few systematic investigations of performance effects, and little consideration of the implications for occupational exposure limits of extensions to the working day. Existing data relate largely to situations where working hours exceed 50 a week and there is a lack of information on hours below this level, which is of direct relevance to the European Community proposal. Finally, it is clear from investigations relating to shiftwork that a range of modifying factors are likely to influence the level and nature of health and performance outcomes. These include the attitudes and motivation of the people concerned, the job requirements, and other aspects of the organisational and cultural climate. It is concluded that there is currently sufficient evidence to raise concerns about the risks to health and safety of long working hours. However, much more work is required to define the level and nature of those risks.