Education and cognitive activity have been suggested to protect against cognitive decline in old age. However, little is known about the long-term effects of extremely high levels of education and intellectual activity across the lifespan. The present study investigated the extent to which these two variables may moderate the age-related differences in cognitive performance in old adults. Therefore, story recall, paired-associates learning, reading span and letter digit performance of 62 university professors (mean age = 72.47) were compared with those of a representative sample of 196 participants of the Zurich Longitudinal Study of Cognitive Aging (mean age = 73.04). The results demonstrate that the highly educated sample performed significantly better than the normally educated sample in the paired-associates learning and reading span test. Furthermore, age effects were found in the letter digit as well as in the paired-associates learning test. While the normally educated sample demonstrated an age-related decrease in the paired-associates learning test, the performance of the highly educated sample actually increased with increasing age. These findings suggest that extremely high levels of education and intellectual activity may postpone age-related deficits in paired-associates learning tasks, but not in speed of processing tasks.