<p>Repetitive head impacts pose a significant threat to brain health. However, their injury accumulation mechanism remains elusive. Here, we study how traumatic axonal injury might accumulate based on typical impact severities and frequencies obtained from University Men’s Ice Hockey. From <i>N</i> = 994 impact simulations over a season (<i>N</i> = 22 athletes), we identified 50th (‘‘mild’’, 7.0% peak fiber strain), 75th (‘‘intermediate’’, 10.6%), and 95th percentile (‘‘moderate’’, 17.9%) peak strain magnitudes in the corpus callosum. We simulated up to <i>N</i> = 11 (95th percentile frequency per player game) stretches sequentially using a male axonal injury model to mimic repeated “mild”, “intermediate”, and “moderate” stretch exposure, along with a moderate insult followed by or following mild/intermediate repetitive stretches. We first identified a “safe” stretch threshold of ~ 5%, below which no cytoskeletal damage or accumulation occurred. Our model predicted that neurofilament (NF) damage accumulated even at the mild magnitude exposure, while damage of tau proteins accumulated at or above the intermediate magnitude. An initial moderate stretch can almost double tau and NF damage accumulation when the axon was exposed to additional intermediate stretches. A history of intermediate stretch exposure prior to a moderate stretch can also increase tau and NF failure (e.g., by ~ 50 and ~ 100%, respectively, relative to a single moderate stretch). These results offer mechanistic insight into why continuing play after concussion would make the brain more vulnerable, and a history of head impacts may lower concussive impact threshold. They may also explain why tau proteins and especially NF-light could serve as concussion-sensitive biomarkers.</p>