It has been established that whereas young children who use their fingers to solve arithmetic problems outperform those who do not, this trend reverses in older children around the age of 7. One possible interpretation is that finger counting is helpful for solving simple problems but becomes inadequate as problem complexity increases, leaving finger users reliant on strategies that are no longer efficient. Another possibility is that nonfinger users after the age of 7 are, in fact, ex-finger users who have already transitioned to efficient mental strategies. These contrasting interpretations carry significant theoretical and educational implications. In the first case, finger counting may be viewed as an ultimately limiting tool, potentially trapping children in immature strategies, and, in the second, as a powerful tool that has not only immediate benefit but also long-term positive effect on children's arithmetic development. To disentangle these interpretations, 192 children from middle to high socioeconomic status (mainly White, 96 girls) were followed from the age of 4.5 to 7.5 across seven testing points. While virtually all children were observed using their fingers at some point during the study, most nonfinger users by age of 6.5 were in fact ex-finger users. Importantly, they presented higher arithmetic performance than finger users and genuine nonfinger users of the same age. These original findings provide the first empirical evidence that finger counting acts as a developmental scaffold toward efficient mental arithmetic, rather than as a mere useful but limiting strategy eventually hindering development. (PsycInfo Database Record (c) 2025 APA, all rights reserved).