In this study, a probabilistic model based on Monte-Carlo theory is applied to predict the fatigue behavior of cast aluminum alloys. The objective of the proposed approach is to investigate the effect of porosity (i.e. the defect size distribution and spatial density) on the fatigue strength and its associated scatter for uniaxial fatigue loads with the load ratio R = 0.1. The proposed model is applied to two cast aluminum alloys with very different defect characteristics. The results for these two alloys confirm that the model can be used to predict the average fatigue strength with a relative error <5%. It also accurately reproduces the experimentally observed trends concerning the scatter in fatigue strength. The scatter is underestimated but is of the same order of magnitude as the experimental values. It is believed that this is because the proposed model considers that the porosity is the only source of scatter. It is demonstrated that the model is well adapted for the prediction of the volume or scale effect in fatigue. This approach can also be used to estimate the Representative Volume Element in Fatigue for cast aluminum alloys.