Between 45-65 million Americans run, making running the most popular physical activity in the US (other than walking). Although running is associated with many health benefits, it also causes as many as 33 running-related injuries (RRIs) per 1000 hours of running. These RRIs cause pain and reduce performance, cause a reduction or cessation of physical activity, necessitate medical visits and expenditures, and cause absenteeism from work, creating a large health and financial burden. Thus, reducing RRI has been a target of much biomechanics research. Unfortunately, this research has not reduced RRI incidence, likely because it has been constrained to lab environments that may misrepresent the number and magnitude of the loads (ground reaction forces) thought to cause RRI. This dissertation overcomes this limitation by using small wearable devices to measure running in realworld settings. First, it presents a first-of-its-kind proof-of-concept study demonstrating that—when measured in the real world—the number and magnitude of loads relates to RRI development. Next, this dissertation presents a series of studies comparing methods to estimate the number and magnitude of loads using small wearable devices then validates these methods across a range of real-world conditions. Finally, applying these validated methods, this dissertation replicates the initial proof-ofconcept study in a large heterogenous sample across a longer period of time. Collectively, these studies demonstrate a novel approach to studying running and RRI, provide the tools to execute that approach, and show support for the hypothesis that the magnitude and number of loads experienced by runners relates to RRI development.