Most circadian texts begin by stating that clocks are pervasive throughout the tree of life. Indeed, clock mechanisms have been described from cyanobacteria to humans, representing a notable example of convergent evolution: yet, there are several phyla in animals, protists or within fungi and bacteria, in which homologs of some-or all-known clock components seem to be absent, posing inevitable questions about the evolution of circadian systems. Moreover, as we move away from model organisms, there are several taxa in which core clock elements can be identified at the genomic levels. However, the functional description of those putative clocks has been hard to achieve, as rhythmicity is not observed unless defined abiotic or nutritional cues are provided. The mechanisms 'conditioning' the functionality of clocks remain uncertain, emphasizing the need to delve further into non-model circadian systems. As the absence of evidence is not evidence of absence, the lack of known core-clock homologs or of observable rhythms in a given organism, cannot be an a priori criterion to discard the presence of a functional clock, as rhythmicity may be limited to yet untested experimental conditions or phenotypes. This article seeks to reflect on these topics, highlighting some of the pressing questions awaiting to be addressed. This article is part of the Theo Murphy meeting issue 'Circadian rhythms in infection and immunity'.