<p>This study presents a multiscale modeling framework to investigate the fatigue behavior of additively manufactured Ti–6Al–4V alloy, with the objective of linking mesoscale cyclic deformation mechanisms to macroscale high cycle fatigue responses. A crystal plasticity-based model is employed to compute fatigue indicator parameters (FIPs), which serve as surrogate measure for fatigue hotspots. The statistical behavior of the highest FIPs is characterized using the Gumbel extreme value distribution, and the distribution parameters are correlated with fatigue life through Bayesian inference following our recent work (Tiwari et al. in Int J Plast 189:104319 <a href="https://doi.org/10.1016/j.ijplas.2025.104319">https://doi.org/10.1016/j.ijplas.2025.104319</a>, 2025). The proposed approach demonstrates strong predictive capability in estimating fatigue life across different loading conditions. Additionally, crack growth behavior is modeled using the Hartman–Schijve equation, and the integration of FIP characteristics into the analysis results in the convergence of crack growth curves for various <i>R</i>-ratios. This convergence highlights the potential of the model to predict crack propagation independently of applied load ratio. It also offers a promising pathway for advancing the understanding of fatigue mechanisms and developing more robust, predictive models for fatigue life and crack growth.</p>