The functions of dance and music in human evolution are a mystery. Current research on the evolution of music has mainly focused on its melodic attribute which would have evolved alongside (proto-)language. Instead, we propose an alternative conceptual framework which focuses on the co-evolution of rhythm and dance (R&D;) as intertwined aspects of a multimodal phenomenon characterized by the unity of action and perception. Reviewing the current literature from this viewpoint we propose the hypothesis that R&D; have co-evolved long before other musical attributes and (proto-)language. Our view is supported by increasing experimental evidence particularly in infants and children: beat is perceived and anticipated already by newborns and rhythm perception depends on body movement. Infants and toddlers spontaneously move to a rhythm irrespective of their cultural background. The impulse to dance may have been prepared by the susceptibility of infants to be soothed by rocking. Conceivable evolutionary functions of R&D; include sexual attraction and transmission of mating signals. Social functions include bonding, synchronization of many individuals, appeasement of hostile individuals, and pre- and extra-verbal communication enabling embodied individual and collective memorizing. In many cultures R&D; are used for entering trance, a base for shamanism and early religions. Individual benefits of R&D; include improvement of body coordination, as well as painkilling, anti-depressive, and anti-boredom effects. Rhythm most likely paved the way for human speech as supported by studies confirming the overlaps between cognitive and neural resources recruited for language and rhythm. In addition, dance encompasses visual and gestural communication. In future studies attention should be paid to which attribute of music is focused on and that the close mutual relation between R&D; is taken into account. The possible evolutionary functions of dance deserve more attention.
