It is presumed and commonly accepted by those skilled in the art of satellite navigation and Kalman filter design that the filter must be provided with the tracker position and velocity a priori in order to determine target position and velocity. Indeed, it is generally asserted that without a priori knowledge (known or measured values) of the tracker position and velocity, line of sight measurements between satellites do not contain adequate information to infer target states. Passive and autonomous navigation of space vehicles without a priori values for the position and velocity of either the target or tracker vehicle is achieved by reconfiguring the extended Kalman filter, or more generally any predictor/correction class filter, to include states for both the target and tracker vehicles. The target and tracker vehicles must both follow trajectories in an inertial frame of reference through the gravitational field of a gravitational body having a known gravitational model. The reconfigured filter simultaneously estimates the position and velocity of both tracking and target space-based vehicles from line-of-sight measurements.