The objective of this paper is to provide a short, updated summary of the neopiagetian models of Pascual-Leone, Case, Fischer, Siegler, Halford et Demetriou. It is then argued that the propositions offered by these models are very close to those of other researchers interested in intelligence, such as researchers in lifespan developmental psychology, in psychometrics or in adult cognition (e.g., Engle or Cowan); in particular, they consider that working memory indexes attentional resources, that develop with age. Finally, two empirical examples are provided. The first one briefly reports results of a longitudinal study in which attentional capacity tasks as well as the Piagetian Balance task were administered to children, initially aged 5 to 10 years. In the second example, working memory tasks were given to children aged 8 to 12 years, together with the Raven's Matrices task. Both studies show that working memory, or attentional. capacity, accounts for most of the age-related variance in the cognitive tasks.