Theoretically, greater motor output variability can inhibit motor learning by inhibiting task acquisition during practice. Although the age-associated differences in motor output variability exacerbate with more difficult tasks, it remains unknown whether task difficulty during task acquisition influences motor learning in older adults. The purpose of this study was to determine whether the difficulty of the practice task affects motor learning in older adults. Twenty four older (72.7±7.4years; 11 women) and 7 young (23.1±4years; 1 man) adults participated in this study. Participants were divided into four groups: 8 older adults who practiced an easy task (O-Easy), 8 older adults who practiced a harder task (O-Hard), 8 older adults who did not practice (O-None), and 7 young adults who did not practice (Y-None). The level of difficulty depended on the relative timing (i.e. phase) of abduction force generation between the index and little fingers to track a moving target on the monitor. The O-Easy group practiced the task with 0°, whereas the O-Hard group practiced the task with 90° relative phase. Practice occurred within a single session for 80 trials. Motor learning was quantified as the ability to transfer the practiced tasks to 45°, 135° and 180° relative phases 24 and 168h after acquisition. Only the O-Easy group was able to significantly transfer the practiced task, as it was indicated by significantly lower force variability and error during all transfer tasks compared with the O-None group (P<0.05). The O-Hard group was not significantly different from the O-None group (P>0.2). In addition, during the transfer tasks the O-Easy group exhibited performance similar to that of the young adults who did not practice. These findings suggest that practice with easier tasks may be advantageous to practice with more difficult tasks to improve motor learning in older adults.