We propose taking event-triggered control actions to implement decentralized control over wireless sensor/actuator networks without requiring synchronized measurement updates. In comparison with the existing results on event-triggered decentralized control, the proposed implementation does not rely on weak coupling between subsystems, nor does it assume the synchronization of local clocks or the existence of a central broadcasting node, and is applicable to nonlinear systems. In addition, higher energy efficiency at the sensors is expected because of the great reduction of the listening times of the sensors. We prove that with asynchronous measurement updates, the event-triggered control actions can guarantee semiglobal practical stability for the sensor/actuator system of interest. We also show that the time between any two consecutive transmissions of measurements at each sensor is bounded from below by a positive constant. Furthermore, asymptotic stability can be achieved when more complicated triggering conditions are introduced. The theoretical analysis is validated by simulations.