This article uses the British Workplace Employee Relations Survey 1998 dataset to evaluate the importance of firm-level factors in determining weekly hours of work in Britain, to analyse the amount of variation of hours within firms, and to assess whether workers are sorted into firms on the basis of hours. Nearly a third of the explained variation in weekly hours of work can be ascribed to firm-level differences, which are unrelated to any observed worker characteristics. Firm-level ‘policies’ are especially important in the private-services sector. A further 40 per cent of variance can be attributed to the effect within firms of different occupation, skills and family characteristics. Finally, about a quarter of the explained variation in working hours is due to a sorting process of workers to firms, especially on the basis of skill and occupation.