The relationship between numerosity discrimination and arithmetic skill reflects the approximate number system and cannot be explained by inhibitory control.

Numerosity discrimination tasks (judging which of two random dotarrays contains the larger number) have been widely used as a measure of the efficiency of an approximate number system (ANS) and are a correlate of early arithmetic skills. Recently, it has been suggested that the relationship between numerosity discrimination and arithmeticis explained by inhibition rather than the ANS. We assessed this hypothesis in astudy of 496 children (mean age = 81.23 months) using numerosity discrimination tasks that manipulated the congruency between surface area and numerosity. Numerosity discrimination for incongruent arrays (which are postulated to require inhibition due to a conflict between judgments based on surface area rather than numerosity) was more difficult than that for congruent arrays. However, all numerosity discrimination tasks showed substantial correlations with each other and correlated with arithmetic. A latent variable path model showed that a general numerosity judgment factorcorrelated witharithmetic even after controlling for a measure of response inhibition. In contrast, numerosity discrimination for incongruent arrays showed no unique relationship with arithmetic ability. Our results do not support the view that the relationship between numerosity discrimination and arithmetic is largely attributable to inhibition; rather, theyare consistent withthe view that numerosity discrimination tasks tap the operation of an ANS.