The aim of the present work was to test the effect of aging on the capacity to perform and validate simple arithmetic calculations. Two experimental groups, eight young and eight old subjects, performed a set of three tasks (A,B,C), requiring different amounts of attentional resources. In test A, the subjects knew in advance the type of operation to be validated; in test B the subjects had to validate items of four randomly mixed operation types; in test C the subjects could use an informative cue which signalled in advance the kind of operation to be performed. The cue was true in 70% of items (valid trials) and false in 30% (invalid trials). The results confirmed a slowing down of cognitive processes in old people that partially depended on task complexity. Differently from young, old subjects could not use the informative cue to orient attention toward specific arithmetic facts: actually in test C the difference between invalid vs. valid trials does not reach statistical significance.