This article synthesizes some of the published literature that selectively compares the cognitive functioning of children with math disabilities (MDs) with average-achieving children and poor readers (children with reading disabilities [RDs] or comorbid disabilities [RDs + MDs]). All studies in the synthesis report reading, IQ, and math scores for children with MDs and poor readers. A random coefficients model of effect sizes (ESs) show that (a) ESs between MD and normal achievers were moderated by variations in working memory and literacy, (b) ESs between MD- and RD-only children were moderated by working memory and problem solving, and (c) ESs between MD and MD + RD children were moderated by long-term memory and IQ scores. No support was found for the notion that the differentiation between MD children and poor readers (RD and MD + RD) was related to variations in reading across the reviewed studies. (Contains 6 tables and 1 note.)