All animals sleep and it is essential for maintaining optimal brain function. However, cetaceans engage in the unusual practice of unihemispherical sleep, where only half of their brain sleeps at a time, due to their constant need for movement and breathing. Most studies of sleep in cetaceans have occurred in captivity. However, tagging devices have now developed to the point where the data collected from wild animals can be assessed against published criteria for defining sleep behaviourally. Seven acoustic and behavioural data loggers were deployed on harbour porpoises, Phocoena phocoena, in Danish waters between May 2010 and August 2011 and stayed on the animals between 53 and 72 h, recording 1884 to 2755 valid dives per animal. Parabolic dives with significantly reduced bioacoustic activity and a stereotyped behavioural pattern were identified as potential sleeping periods. The recordings for nearly half of the parabolic dives were found to contain no vocalization (echolocation clicks), significantly more than other dive types. Of the remaining parabolic dives, the click rate was, when normalized to their individual means, also significantly lower than detected in other dive types. Additionally, parabolic dives were shallow compared to all other dive types, and found to have a stereotypic low-energy profile. They were also found to contain fewer rolls and incorporate a lower vertical descent rate than most other dive types. If the data are representative, harbour porpoises spend a small, but meaningful amount of their diving time engaged in parabolic dives and thus potentially sleeping. All animals have a fundamental need for undisturbed sleep. These quiet periods thus need to be considered in studies of anthropogenic effects, but also those employing passive acoustic monitoring techniques, as well as in efforts to reduce incidental bycatch in fisheries, given the associated periods of reduced environmental awareness.