Abstract: A method for selecting a set of four target bearings from a plurality of bearing measurements to optimize rapidity, accuracy and stability of a target track solution in a bearings-only target motion algorithm. Four bearings are selected to generate the deterministic solution by first selecting a candidate bearing set, then computing a set of "n" solutions from the candidate set and others adjacent thereto. Motion parameters are then computed, and any solution exhibiting parameters outside a user defined deviation from the mean is discarded. The mean target parameters of the remaining solutions may again be computed, and further culling out performed, until the desired distribution is achieved. An optimal solution is chosen as the solution from the remaining sample space that is closest to the mean in target range, course and speed.