This paper considers the design of a Target Tracking Wireless Sensor Networks (WSN). Target tracking refers to the process of detecting the presence of a target and locating it in the field with collaborative processing. Since the accuracy of tracking the target is dependent upon the number of active sensor nodes at a given time instant, it is possible to reduce the energy consumed by reducing the number of actively tracking sensor nodes. The success of such a scheme will depend upon the application's tolerance to tracking error. This paper proposes a target tracking methodology in which the application specified tracking error tolerance is used to activate only a subset of nodes by predicting the target's future trajectory. The proposed method is compared with existing activation schemes using an OMNeT based discrete-event simulation model, in terms of the energy vs. error tolerance trade-off. The proposed approach is shown to outperform most of the existing schemes in several scenarios. © 2011 IEEE.