[eng] Summary Rest in poikilothermic animals is an adaptation of the organism to adjust to the geophysical cycles, a doubtless valuable function for all animals. In this review, we argue that the function of sleep could be trivial for mammals and birds because sleep does not provide additional advantages over simple rest. This conclusion can be reached by using the null hypothesis and parsimony arguments. First, we develop some theoretical and empirical considerations supporting the absence of specific effects after sleep deprivation. Then, we question the adaptive value of sleep traits by using non-coding DNA as a metaphor that shows that the complexity in the design is not a definitive proof of adaptation. We then propose that few, if any, phenotypic selectable traits do exist in sleep. Instead, the selection of efficient waking has been the major determinant of the most significant aspects in sleep structure. In addition, we suggest that the regulation of sleep is only a mechanism to enforce rest, a state that was challenged after the development of homeothermy. As a general conclusion, there is no direct answer to the problem of why we sleep; only an explanation of why such a complex set of mechanisms is used to perform what seems to be a simple function. This explanation should be reached by following the evolution of wakefulness rather than that of sleep. Sleep could have additional functions secondarily added to the trivial one, although, in this case, the necessity and sufficiency of these sleep functions should be demonstrated.