Earthquakes have great impact on civil infrastructure, and the consequence of these events can be catastrophic. Not only are such damaging events rare, but they are unpredictable. Therefore, detection and early mitigation of damage is a critical issue to ensuring safety of civil infrastructure under earthquakes. One of the challenges to earthquake monitoring systems for structures is cost. Wireless sensors offer tremendous opportunity to reduce cost and realize the dream of pervasive sensing. However, earthquake monitoring of civil infrastructure using wireless smart sensors (WSS) is challenging. WSSs typically employ some form of duty cycling to reduce power consumption; hence, they will miss earthquakes if the sensor is in sleep mode when they occur. Demand-based WSSs meet the requirements of always-on earthquake monitoring within a minimal power budget by integrating an ultra-low power trigger accelerometer with a high-fidelity WSS platform and combining the beneficial aspects of both. In particular, the approach is able to rapidly turn on the WSS upon the occurrence of earthquakes and seamlessly transition to high-fidelity data acquisition. Afterwards, post-sensing data fusion is conducted to integrate the data obtained from the two sensors without losing any critical data from the triggering event. A laboratory test is presented to verify the efficacy of the proposed idea. The results show that the demand-based WSSs are able to capture the occurrence of earthquake and provide high-fidelity data for structural condition assessment in a timely manner.