In this thesis, we present the design and evaluation of the SPOT power metering system, and its application in run-time energy management. SPOT is a scalable power observation tool that enables in situ measurement of nodal power and energy over a dynamic range exceeding four decades and at a temporal resolution of microseconds. Using SPOT, every node in a sensor network can be instrumented, providing unparalleled visibility into the dynamic power prole of applications and system software. Power metering at every node enables previously impossible empirical evaluation of low power designs at scale. The SPOT architecture and design meet challenges unique to wireless sensor networks and other low power systems, such as orders of magnitude difference in current draws between sleep and active states, short-duration power spikes during periods of brief activity, and the need for minimum perturbation of the system under observation. Using SPOT as an underlying service, we present a run-time energy management architecture (EMA), closing the feedback loop between observation and actuation. EMA adopts a three component decomposition – a policy interface for user input, a mechanism to monitor system and component resource usage, and a management module for enforc-ing policy directives. EMA provides facilities for 1) individual accountability of energy-consuming units, 2) graceful degradation and predicable operation in a dynamic environment, and 3) expression of network-level policies that can be used by individual nodes for local optimization and shared among nodes to adjust behavior of the network as a whole.