Ice hockey puck impacts combine high velocity, oblique loading, and spatially variable contact, yet conventional helmet assessments rely primarily on global head kinematics. This study quantified how puck impact velocity, angle, and location affect skull stress and brain tissue strain. A commercially available ice hockey helmet was reconstructed from three-dimensional scans and coupled with the Total Human Model for Safety (THUMS) Version 7.1 head model. Helmet response was benchmarked against standardized drop-test data at top, lateral, and rear locations. Eighteen nominal puck-impact conditions combined two velocities (30 and 40 m/s), three angles (0°, 45°, and 60°), and three locations; additional ±10% velocity runs assessed sensitivity. Peak skull von Mises stress (skull stress) and brain maximum principal strain (MPS) were compared descriptively. Benchmark peak linear acceleration differed from reference data by 0.5–5.8%. The greatest skull stress occurred at 40 m/s, 0°, lateral impact (0.556 MPa), whereas the greatest MPS occurred at 40 m/s, 0°, top impact (0.175). These different worst-case locations show that local skull loading and brain tissue deformation are not captured by a single metric. Multi-metric, location-specific evaluation may improve ice hockey helmet assessment and regional optimization.