Young children can quickly and intuitively represent the number of objects in a visual scene through the Approximate Number System (ANS). The precision of the ANS - indexed as the most difficult ratio of two numbers that children can reliably discriminate - is well known to improve with development: whereas infants require relatively large ratios to discriminate number, children can discriminate finer and finer changes in number between toddlerhood and early adulthood. Which factors drive the developmental improvements in ANS precision? Here, we investigate the influence of four non-numeric dimensions - area, density, line length, and time - on ANS development, exploring the degree to which the ANS develops independently from these other dimensions, from inhibitory control, and from domain-general factors such as attention and working memory that are shared between these tasks. A sample of 185 children between the ages of 2 and 12 years completed five discrimination tasks: approximate number, area, density, length, and time. We report three main findings. First, logistic growth models applied to both accuracy and Weber fractions (w; an index of ANS precision) across age reveal distinct developmental trajectories across the five dimensions: while area and length develop by adolescence, time and density do not develop fully until early adulthood, with ANS precision developing at an intermediate rate. Second, we find that ANS precision develops independently of the other four dimensions, which in turn develop independently of the ANS. Third, we find that ANS precision also develops independently from individual differences in inhibitory control (indexed as the difference in accuracy and w between Congruent and Incongruent ANS trials). Together, these results are the first to provide evidence for domain-specific improvements in ANS precision, and place children's maturing perception of number, space, and time into a broader developmental context.