The problem of tracking mobile users in Personal Communication Service (PCS) networks is discussed. We propose a novel approach for reducing the wireless cost of tracking users. The basic idea is to use non-utilized system resources for initiating queries about the location of mobile users, in addition to the process of user registration. Queries are applied at each cell, independently of the other cells, whenever the load on the local control channel drops below a pre-defined threshold. Our study focuses on two issues: (1) proposing the initiated queries approach and an algorithm for its application, and (2) studying and quantifying the value of location information and evaluating the parameters affecting it. Our analysis shows that the expected benefit due to location knowledge in a Markovian motion model depends, among other things, on the determinant of the transition matrix and on the variability of the location distribution function. The active tracking approach, as opposed to other dynamic strategies, does not require any modification of user equipment. The importance of this property is in its practicality: An implementation of a new registration strategy in current systems would require a modification of the users equipment. Moreover, the proposed method can be easily implemented in addition to any known tracking strategy, to reduce further the tracking cost. The performance of the active tracking method is evaluated under two registration strategies: The geographic-based strategy, currently used in cellular networks, and the profile-based strategy, suggested elsewhere. Under both strategies, it significantly reduces the tracking cost.