Research suggests that allowing individuals to control their own practice schedule has a positive effect on motor learning. In this experiment we examined the effect of task difficulty and self-regulated practice strategies on motor learning. The task was to move a mouse-operated cursor through pattern arrays that differed in two levels of difficulty. Participants learned either four easy or hard patterns after assignment to one of four groups that ordered practice in blocked, random, self-regulated, and yoked-to-self-regulated schedules. Although self-regulation provided no special benefit in acquisition, these groups showed the most improved performance in retention, irrespective of task difficulty. Although individual switch strategies for members of the self-regulated groups were quite variable, the impact of self-regulation on motor learning remained similar. These findings add to the growing body of literature suggesting that self-regulated practice is an important variable for motor learning. (Contains 4 figures.)