The present research examined if the time needed to implement expectancy-based strategic processes is different in younger and healthy older adults. In four experiments participants from both age groups performed different strategic priming tasks. These included a greater proportion of incongruent (or unrelated; 80%) than of congruent (or related; 20%) trials. With this procedure performance is worse for congruent (less frequent) than for incongruent (more frequent) trials, thus demonstrating that the relative frequency information can be used to predict the upcoming target. To explore the time course of these expectancy-based effects, the prime-target SOA was manipulated across experiments through a range of intervals: 400, 1000 and 2000 ms. Participants also performed a change localization and an antisaccade task to assess their working memory and attention control capacities. The results showed that increases in age were associated with (a) a slower processing-speed, (b) a decline in WM capacity, and (c) a decreased capacity for attentional control. The latter was evidenced by a disproportionate deterioration of performance in the antisaccade trials compared to the prosaccade ones in the older group. Results from the priming tasks showed a delay in the implementation of expectancies in older adults. Whereas younger participants showed strategic effects already at 1000 ms, older participants consistently failed to show expectancy-based priming during the same interval. Importantly, these effects appeared later at 2000 ms, being similar in magnitude to those by the younger participants and unaffected by task practice. The present findings demonstrate that the ability to implement expectancy-based strategies is slowed down in normal aging.