A measurement-to-measurement data association problem is formulated for a target tracking system consisting of one or two 2D sensors and a 3D sensor. Operating conditions are identified under which performance is improved by using two 2D sensors and a 3D sensor instead of one 2D sensor and a 3D sensor. To facilitate this study, two algorithms are introduced to compute near-optimal solutions of the corresponding three-way assignment problem: a single-step algorithm based on two independent two-way assignment problems, and a related iterative algorithm that explicitly enforces a compatibility condition between measurements made by the 2D sensors. Simulation studies show that the position estimates obtained with the three-sensor system are much more accurate than those obtained with a two-sensor system whenever there is large uncertainty in the 3D sensor in the dimension orthogonal to the plane of the 2D sensor in the two-sensor system. Moreover, whenever there is large uncertainty in the measurements from the 3D sensor in the common dimension of the 2D sensors, the percentage of correct matches with the the iterative assignment algorithm for the three-sensor system is significantly better than that with a two-sensor system. The degree to which the methods and results can be extended to more realistic 3D radar and 2D camera models is discussed and inferences for aerospace and missile defense applications are drawn.