The spring dawn and dusk chorus of birds is a widespread phenomenon, yet its origin remains puzzling. We propose that a dawn and dusk chorus will inevitably arise if two criteria are met: (1) females leave their roost later in the morning and go to roost earlier in the evening than their mate, and (2) males sing more when separated from their mate. Previous studies on blue tits (
Cyanistes caeruleus
) support the first criterion. We here report that males sing at a higher rate whenever they are separated from their mate and that song rate increases with the duration of female absence. These findings can explain the existence of the dawn and dusk chorus in blue tits, and they can explain why the dawn chorus is more pronounced than the dusk chorus, as is typically observed. An exhaustive literature search provides support for both criteria of the ‘absent mate’ hypothesis in several passerine birds. We found no evidence contradicting the hypothesis. The new hypothesis is not inconsistent with many of the existing hypotheses about dawn singing, but may be a more general explanation for the occurrence of a dawn and dusk chorus. We describe how the ‘absent mate’ hypothesis leads to testable predictions about daily and seasonal variation in song output.