||An interest in wearable sensors for activity tracking and health monitoring, coupled with the desire for energy-independent devices for increased convenience and an improved user experience, have resulted in significant research and commercial interest in wrist-worn energy harvesting. Energy harvesting - an alternative to finite, traditional energy storage technologies, such as batteries - has the potential to provide power to wearable devices for a functionally unlimited amount of time by extracting ambient, freely-available energy from the environment, obviating the need for user intervention in replenishing a finite energy supply. Vibration energy harvesting, which concerns the harvest of kinetic energy, is of particular interest for wearable applications. However, the low-frequency, highamplitude excitations that typify wrist motion during common daily activities make vibration energy harvesting using traditional linear vibration energy harvesting architectures challenging. Alternatives to linear vibration energy harvesting architectures have been proposed for wrist-worn energy harvesting, with eccentric rotor harvesters - asymmetric, rotational devices - representing a common choice in the literature (possibly due to the watch-like form factor and extant commercial products) that exhibit some interesting properties. This project concerns the design, analysis, and optimization of architectures for wrist-worn energy harvesting, with particular emphasis placed on eccentric rotor harvesting architectures. The problem of determining which architecture is best suited for iv wrist-worn energy harvesting is first approached by deriving mathematical models of several common architectures and developing a means by which the disparate architectures may be compared fairly. Eccentric rotor harvesters - especially those that include a torsional spring - fare particularly well, and become the focus of the remainder of the work. Several generations of eccentric rotor prototypes are designed, fabricated, and characterized in order to validate the mathematical models and demonstrate the improvement of power output that comes with the addition of a torsional spring to the eccentric rotor architecture. Finally, a dynamical analysis gives insight into how the design parameters affect power output and provides an explanation for some of the nonlinear behavior observed in these devices. This knowledge is used to develop a new kind of eccentric rotor harvester that may have significant advantages over designs hitherto presented in the energy harvesting literature.