Microstructure plays a fundamental role in the fatigue performance of metallic alloys, as microstructural cracks typically nucleate in regions where microscopic plastic strain localizes. Microstructure sensitive fatigue life prediction models rely on the simulation of the cyclic response a polycrystal using computational homogenization, including the microstructure details in the Representative Volume Element, to obtain Fatigue Indicator Parameters (FIPs) which correlate with fatigue life. The resulting models provide good predictions but usually limited to a particular cycle range, and models covering from LCF to the fatigue limit are uncommon. The objective of this work is to propose and validate an statistical model for fatigue life prediction in AL7075 valid for the full cycle range for a stress-based loading condition with a reverse ratio of R = -1. The approach is based on the combination of a plastic FIP (based on energy dissipation), adequate for LCF conditions with large plasticity, with an elastic FIP, more suitable for HCF where microplasticity is very limited. Theelastic FIP used is ibased on the Crossland criterion, which combines resolved stress on octahedral planes with hydrostatic stress. The results show the approach proposed successfully captures the experimental results throughout the full S-N curve, from 10^4 cycles to the fatigue limit, where fatigue life is over 10^7 cycles.