Porosity is one of the most common defects to degrade the mechanical properties of aluminum alloys. Prediction of pore size, therefore, is critical to optimize the quality of castings. Moreover, to the design engineer, knowledge of the inherent pore population in a casting is essential to avoid potential fatigue failure of the component. In this work, the size distribution of the porosity was modeled based on the assumptions that the hydrogen pores are nucleated heterogeneously and that the nucleation site distribution is a Gaussian function of hydrogen supersaturation in the melt. The pore growth is simulated as a hydrogen-diffusion-controlled process, which is driven by the hydrogen concentration gradient at the pore liquid interface. Directionally solidified A356 (Al-7Si-0.3Mg) alloy castings were used to evaluate the predictive capability of the proposed model. The cast pore volume fraction and size distributions were measured using X-ray microtomography (XMT). Comparison of the experimental and simulation results showed that good agreement could be obtained in terms of both porosity fraction and size distribution. The model can effectively evaluate the effect of hydrogen content, heterogeneous pore nucleation population, cooling conditions, and degassing time on microporosity formation.