Energy usage evaluation and condition monitoring for electric machines are important in industry for overall energy savings. Traditionally these functions are realized only for large motors in wired systems formed by communication cables and various types of sensors. The unique characteristics of the wireless sensor networks (WSN) make them the ideal wireless structure for low-cost energy management in industrial plants. This work focuses on developing nonintrusive motor-efficiency-estimation methods, which are essential in the wireless motor-energy-management systems in a WSN architecture that is capable of improving overall energy savings in U.S. industry. This work starts with an investigation of existing motor-efficiency-evaluation methods. Based on the findings, a general approach of developing nonintrusive efficiency-estimation methods is proposed, incorporating sensorless rotor-speed detection, stator-resistance estimation, and loss estimation techniques. Following this approach, two new methods are proposed for estimating the efficiencies of in-service induction motors, using air-gap torque estimation and a modified induction motor equivalent circuit, respectively. The experimental results show that both methods achieve accurate efficiency estimates within ¡À2-3% errors under normal load conditions, using only a few cycles of input voltages and currents. The analytical results obtained from error analysis agree well with the experimental results. Using the proposed efficiency-estimation methods, a closed-loop motor-energy-management scheme for industrial plants with a WSN architecture is proposed. Besides the energy-usage-evaluation algorithms, this scheme also incorporates various sensorless current-based motor-condition-monitoring algorithms. A uniform data interface is defined to seamlessly integrate these energy-evaluation and condition-monitoring algorithms. Prototype wireless sensor devices are designed and implemented to satisfy the specific needs of motor energy management. A WSN test bed is implemented. The applicability of the proposed scheme is validated from the experimental results using multiple motors with different physical configurations under various load conditions. To demonstrate the validity of the measured and estimated motor efficiencies in the experiments presented in this work, an in-depth error analysis on motor efficiency measurement and estimation is conducted, using maximum error estimation, worst-case error estimation, and realistic error estimation techniques. The conclusions, contributions, and recommendations are summarized at the end.