Since the 1980s, working hours have tended to become more diverse and flexible. These developments impact not only the possibilities of synchronising social time but also pay levels. Duration indicators become less relevant, since they disconnect the time worked and paid from the impact of the work on employees’ lives. The objective of this paper is to analyse the effects of this disconnection by measuring the fragmentation of the working day. We propose an indicator of ‘density of daily working time’ which links two specific durations (the extent of the working day and the average daily duration of working). This indicator shows that a low density of the working day makes it possible to circumvent the minimum wage principle, particularly in the case of lower-skilled service occupations. These results raise questions about how working time is measured and about its regulation.