Students who complete kindergarten with an inadequate knowledge of basic mathematics concepts and skills will continue to experience difficulties with mathematics throughout their elementary and secondary years and may be at increased risk for math disabilities. There is a critical need to identify students experiencing difficulties in mathematics in the early elementary grades and to provide immediate and targeted instruction to remediate these deficits. Most early math screening tools focus on only a single skill, resulting in an incomplete picture of student performance and limited predictive validity. To address this need, we are developing a multiple-gating system of math assessment, the Primary Math Assessment (PMA), that both screens and provides diagnostic information in six domains. In this study, we present the results of the development and validation of items across the domains that will comprise the PMA. Multidimensional Rasch models were used to estimate theoretically plausible dimensionality structures. Parsimony fit indices supported the six-dimensional model as the most generalizability model for the PMA data and supports reporting of six separate scores.