This paper presents a reachability set based method for tracking maneuvering space objects in the presence of sparse measurements. The proposed approach invokes the Dynamic Data Driven Application Systems (DDDAS) paradigm by dynamically integrating model forecasts due to uncertainties in maneuver capabilities with collected sensor data to search and track for a non-cooperative satellite in a control theoretic framework. The typically large time interval between measurements from ground stations presents significant problems for tracking satellites that have maneuvered during this interval. Using reachability set propagation techniques and a particle filter update scheme, an intelligently guided search algorithm is developed. This algorithm enables the systematic reduction of likely reachable states until measurements of the target are acquired and traditional tracking techniques can be resumed. Numerical simulations of a space-based sensor tasking scenario are given, however, the method is generic and can be extended to ground-based sensors or a combination of both ground and space-based sensors.