<p>A statistical fatigue-based process window model was developed for Ti-6Al-4V laser powder bed fusion (PBF-LB) additive manufacturing (AM). The process window model is a six-dimensional surface, screened from over 300 recorded process and fatigue testing variables. It was developed by fusing process data across multiple PBF-LB machine models and operating facilities with differing test plans, while maintaining statistical confidence regarding the coherence of pooled data for its use in this modeling setting. The model yields continuous predictions of failure probabilities and uncertainties and highlights families of process setting which are all expected to meet minimum specified property requirements despite changing process variables. Simultaneous process variable confidence intervals and factors of safety can be established at single values and across process families. Additionally, the impact of variable perturbations can be analyzed regarding expected life and life uncertainty. This work illustrates that it is possible to analyze and compare continuous families of PBF-LB processes, despite building models from federated data, while maintaining the confidence typically associated with the “point design” approaches common to AM industrial practice. This work also illustrates how a single process model can represent multiple machines and facilities for a specific material, while still enabling model users to compare each part against a common baseline.</p>