A new computational model for predicting microporosity in aluminum alloys is described. The model was calibrated against literature data for binary Al-7 pct Si alloys, and subsequently applied to a chill plate test casting in A356 alloy and to an engine block in 319 alloy. The new model allows spherical micropores to nucleate and grow by hydrogen diffusion from a material volume surrounding the pores. This differs from a conventional interdendritic flow computational model for calculating porosity that assumes spherical pores have a diameter proportional to the secondary dendrite arm spacing (SDAS). The new integrated pore growth and interdendritic flow model predicts larger pore diameters and a volume fraction of microporosity that is in better agreement with experimental observations than the interdendritic flow model.