In 'real life,' we learn motor tasks as integrated units, consisting of the motor, spatial, and temporal elements of each task; (1-4) in laboratory investigations, however, the elements of skilled [...], Background and Purpose The purpose of this study was to identify which characteristics of a multidimensional sequence containing motor, spatial, and temporal elements would be most salient for motor sequence learning and whether age might differentially affect this learning. Subjects Younger (n = 11, mean age = 26.0 years), middle-aged (n = 13, mean age = 50.7 years), and older (n = 11, mean age = 77.5 years) adults who were neurologically intact participated in the study. Methods Participants practiced a sequencing task with repeated motor, spatial, and temporal dimensions for 2 days; on a separate third day, participants completed retention and interference tests designed to assess sequence learning and which elements of the sequence were learned. The mean median response time for each block of responses was used to assess motor sequence learning. Results Younger and middle-aged adults demonstrated sequence-specific motor learning at retention testing via faster response times for repeated sequences than random sequences; both of these groups showed interference for the motor dimension. In contrast, older adults demonstrated nonspecific learning (ie, similar improvements in response time for both random and repeated sequences). These findings were shown by a lack of difference between random and repeated sequence performance in the older adult group both at retention testing and during interference tests. Conclusion and Discussion Our data suggest that, when younger and middle-aged adults practice sequences containing multiple dimensions of movement, the motor element is most important for motor learning. The absence of sequence-specific change demonstrated by an older adult group that was healthy suggests an age-related impairment in motor learning that may have profound implications for rehabilitation.