The goal of the current study was to explore learning and short-term retention using a modified serial reaction time task. The multi-finger sequence task was designed to present repeated and random sequences in a completely interleaved fashion, giving participants within block, variable practice, on the two types of sequences. Eighteen younger adults (M(age) = 24 years) and 15 older adults (M(age) = 65 years) participated in the experiment. Participants were asked to respond on a piano keyboard to a visual stimulus that appeared in one of four squares on the computer screen. They were not informed that one of the sequences presented would repeat. Sequence-specific learning, within-day and across-days, was inferred from differences in accuracy and reaction time between repeated and random sequences. Age equivalence was observed in sequence-specific learning and retention across days, and suggests that older adults may benefit from variable practice.