In the ultimate quest to achieve predictive capabilities for crack evolution across multiple length scales, the final generation of Prof. Ingraffea’s graduate students stood on the shoulders of their predecessors, leveraging some of the most advanced materials-characterization and modeling techniques to capture crack geometries and environments with utmost fidelity. The first part of this talk highlights two novel, numerical toolsets that were developed to enable the prediction of 3D crack propagation at the structural or component length scale. The first toolset is one that uses material-state mapping along with FRANC3D-inspired adaptive remeshing to predict propagation of 3D cracks in ductile materials. The second toolset was developed to predict 3D crack-shape evolution by calculating local increments of crack extension, Δai, using energy-release-rate principles. The second part of the talk highlights novel characterization and modeling efforts that were carried out to understand (and eventually to predict) the formation and early propagation of 3D cracks at the microstructural length scale. In one effort, 3D characterization of fatigue-crack nucleation in a Ni-base superalloy microstructure was reconstructed using 3D crystal-plastic finite-element (CPFE) modeling. “Big data” concepts were utilized to discover quantitative correlations between the underlying microstructure and fatigue indicator parameters computed from the CPFE results. In another effort, the propagation of a microstructurally small fatigue crack in an aluminum alloy was characterized in 3D for the first time. The 3D measurements were converted to a 3D CPFE model that explicitly represented the history-dependent shape of the 3D fatigue crack as well as the surrounding grain structure. The talk concludes with important lessons learned in the Cornell Fracture Group and a look to the future.