The parallel training-related BOLD increases for letter memory and 3-back were not seen in the old. Thus, the pattern that emerges from these data suggests that one prerequisite for transfer of learning is that the criterion and transfer tasks engage similar brain systems, and that the tasks are similarly responsive to changes in these systems as a function of training (Jonides, 2004). Further, the data point to a critical role of the striatal complex in WM improvement and transfer. Toward this end, marked age-related losses in striatal morphology (Raz et al., 2003; Raz et al., 2005) and neurotransmitter efficacy (Bäckman et al., 2006, 2010) might constrain older adults regarding the magnitude and generalizability of gains from WM training.