Daylight saving time is currently adopted in over 70 countries and imposes a twice yearly 1 h change in local clock time. Relative ease in adjustment of sleep patterns is assumed by the general population but this review suggests that the scientific data challenge a popular understanding of the clock change periods. The start of daylight saving time in the spring is thought to lead to the relatively inconsequential loss of 1 h of sleep on the night of the transition, but data suggests that increased sleep fragmentation and sleep latency present a cumulative effect of sleep loss, at least across the following week, perhaps longer. The autumn transition is often popularised as a gain of 1 h of sleep but there is little evidence of extra sleep on that night. The cumulative effect of five consecutive days of earlier rise times following the autumn change again suggests a net loss of sleep across the week. Indirect evidence of an increase in traffic accident rates, and change in health and regulatory behaviours which may be related to sleep disruption suggest that adjustment to daylight saving time is neither immediate nor without consequence.