Arithmetic abilities are required when solving problems such as “7×8” or “24 + 33” and are frequently used in everyday life situations. A dominant theoretical view holds that our arithmetic abilities develop from an approximate number system (ANS), considered to be robust to ageing. Consistent with this idea, recent work demonstrated that arithmetical abilities are also resilient to ageing. On the other hand, arithmetic abilities also rely on diverse general cognitive processing mechanisms (e.g. rote verbal memory, sensitivity to interference, updating and shifting) which tend to decline with age. On the basis of this view, arithmetic abilities would be expected to decline with age. Currently, the question of whether and how arithmetic abilities are affected by the decline of cognitive processing remains largely unanswered. The goal of our research is to fill in this gap and to investigate whether and how arithmetic abilities change with ageing. Are arithmetic abilities resilient to ageing such as the ANS theory would predict? Or, are they declining with age, along with the decline observed for other related general cognitive mechanisms? Different from other previous studies, we will adapt our arithmetic tasks so that they directly and intrinsically tap on the cognitive mechanisms in question. At the theoretical level, this project will allow to demonstrate whether arithmetic abilities depend more on the ANS or on general cognitive processing mechanisms. More precisely, our approach allows to specify which arithmetical process is related to what specific cognitive process.