It has been suggested that a perceived rhythmic organization may mediate remembering musical and other stimuli. This experiment examined whether or not Ss could register and remember position in a sequence using rhythmic grouping. Sixteen Ss heard tapped sequences 24-63 beats long, accented to encourage grouping by 4s. Ss tapped responses revealing whether or not they remembered the sequence length-the last tap's position. Significantly more incorrect responses were off by multiples of 4 beats than by adjacent amounts; these frequent errors of whole rhythmic groups of 4 showed that Ss coded sequence length rhythmically. Ss proved 53% accurate over four response conditions, with individual's scores ranging from 8% to 87%. It was concluded that Ss could count with rhythmic hierarchies, essentially equivalent to counting with a nonstandard number base, to code sequential position.