Running is one of the most popular forms of physical activity, however about half of runners are injured each year. Patellofemoral pain (PFP) is one of the most common running related injuries and is associated with several biomechanical factors. Yet, prior research suggests PFP subgroups with unique factors exist. Identifying, characterizing, and classifying subgroups could provide important information for developing targeted treatments that may be more effective at mitigating PFP. Purpose:​ The purposes of this study were to 1) identify and characterize PFP subgroups with kinematic and kinetic variables from pain exacerbating activities, 2) classify subgroups with cost-effective and user-friendly wearable sensors and supervised machine learning models, and 3) investigate the impact of a prolonged running bout on biomechanical characteristics of subgroups to determine if activity may have important implications for subgrouping. Methods:​ 40 individuals with a diagnosis of PFP and 20 healthy controls performed running, squatting, lunging, and jumping trials in a motion capture environment. Kinematic and kinetic features were fed into a clustering model, and characteristics of each group were quantified and compared. Subgroups were then classified with inertial measurement unit data and a hierarchical support vector machine (SVM). Effects of the impact of a prolonged run and development of pain on subgroups were also assessed. Results:​ Four subgroups, three of which contained mostly individuals with PFP (C1, C3, C4), were identified. The first subgroup (C1) was characterized by excessive knee abduction and rotation moments, and reduced rearfoot inversion moments. The third subgroup (C3) was characterized by greater frontal plane loading rates and greater rearfoot eversion range of motion. The fourth subgroup (C4) displayed proximal kinematic adaptations and greater knee abduction angles. Subgroup classification accuracy at child nodes of a hierarchical support vector machine were 81.37% and 93.54%, respectively. The prolonged running bout did not appear to affect identified subgroups. Conclusions:​ This was the first study to identify subgroups with biomechanical variables during pain exacerbating activities and classify subgroups with wearables. Clinicians could identify a patient’s subgroup in a clinical setting with this approach to help design and implement more targeted treatment approaches.