||Quantification of joint angles and translations (i.e., kinematics) is essential to understand how injury and trauma affect joint mechanics. Despite the large number of injuries, there is limited research regarding the mechanics of the joints in the foot or ankle. To address this dearth of knowledge, the overarching goal of this dissertation was to develop, validate, and apply dynamic imaging to quantify in vivo kinematics of the foot and ankle. The first theme of this dissertation detailed the development and validation of dual fluoroscopy (DF) and model-based tracking to quantify three dimensional (3D) kinematics of the ankle. A cadaver study was performed to evaluate the accuracy of the DF system against the reference standard, dynamic radiostereometric analysis. The mean angular and translational biases of the DF system did not exceed 1.7° and 0.5 mm, respectively. The validated DF system was used in the second theme of this dissertation to assess and compare tibiotalar and subtalar kinematics in control subjects and patients with chronic ankle instability (CAI). In controls, the tibiotalar joint was found to provide significantly more dorsi/plantarflexion range of motion (ROM) than the subtalar joint. Conversely, the subtalar joint provided significantly more inversion/eversion ROM than the tibiotalar joint. During heelrise, controls exhibited increased tibiotalar and subtalar internal rotation and subtalar inversion going into plantarflexion, while CAI patients exhibited increased tibiotalar and subtalar external rotation and subtalar eversion. These iv altered mechanics may indicate potential characteristics of CAI. Due to several limitations, DF is impractical for many clinical and research applications. As such, the third theme of this dissertation evaluated the accuracy of the modified Shriners Hospitals for Children Greenville skin marker model, using DF as the reference standard. Significant differences between skin markers and DF were often seen in ankle and midfoot angles during large ROM and the loading and unloading portions of stance, indicating that skin marker models may be less accurate during highly dynamic activities. These data will be used to improve model accuracy and provide improved tools for clinical and research applications.