Two meta-analyses assessed whether the relations between reading and mathematics outcomes could be explained through overlapping skills (e.g., systems for word and fact retrieval) or domain-general influences (e.g., top-down attentional control). The first (378 studies, 1,282,796 participants) included weighted random-effects meta-regression models to explore and contrast the magnitudes of the links between different reading and mathematical competencies. The second (138 studies, 39,836 participants) used meta-analytic structural equation modeling to determine the influence of a domain-general factor, defined by intelligence, executive functioning, working and short-term memory, and processing speed measures, on the link between reading and mathematics skills. The overall relation was significant (r=0.52), as were all associations between specific reading and mathematics measures (rs = 0.23 to 0.61, ps<.05). Most of the correlations were similar across different types of reading and mathematics competencies, although generally smaller than within-domain correlations. The domain-general model explained most of the covariance between reading and mathematics outcomes, with a few modest moderating effects (e.g., age). The results imply correlations between reading and mathematics measures are largely due to domain-general processes, although within-domain correlations confirm the importance of overlapping competencies especially for reading.