This work constitutes an effort to examine a new, low-cost orbit determination method for Earth-orbiting satellites which transmit an always-on telemetry signal. Currently, orbit determination for many commercial satellites is performed using two-way ranging via 10-15 m high-gain antennas. The cost of orbit determination is driven largely by the capital equipment costs of purchasing the necessary earth stations or leasing time on existing stations. For a typical orbit-raising campaign, the cost to lease earth station time may exceed $100000. The new orbit determination method employs only low-cost, low-gain antennas and general purpose computing equipment. The cost to purchase an entire network of low-cost ground stations, which could be employed to perform orbit determination for many orbit raising campaigns, should be commensurate with the cost of leasing earth station time for a single orbit raising campaign. The concept of operations of the new, low-cost method includes several ground stations, probably not less than 10 and not more than 20, each equipped with a low-gain antenna for receiving the Ku band spacecraft telemetry signal, down-conversion equipment, digitization equipment, and a general-purpose computer. Each ground station would observe a spacecraft's normal telemetry signal at intervals and derive one or more signal observables to be used in an orbit determination filter or other orbit estimation algorithm. This operational concept differs from the currently-employed concepts in that it does not entail the enormous capital equipment costs that accompany typical two-way ranging techniques. A potential drawback of this method is that since it employs only low-gain antennas, the signal-to-noise ratio is necessarily lower, and the downlink telemetry bit rate is concomitantly lower. This dissertation provides a signal model for the specific signal used for testing, as well as coarse and fine signal detection algorithms. These algorithms are tested on data gathered from an on-orbit satellite signal. The fine acquisition algorithm provides a method of deriving signal observables usable by an orbit determination algorithm. Finally, an analysis of orbitdetermination performance using only the most simple observable that can be derived from the signal detection algorithm is presented. For a modest number of ground stations, as few as six, real-time orbit-determination accuracy on the order of a kilometer can be expected. Even in the worst cases, simulated position estimation errors climb to only a few kilometers RMS. When more ground stations are employed, the simulated orbit determination accuracy reaches a few tens of meters. This method, or a derivative of it, holds great potential for providing medium-precision orbit determination for a small fraction of the cost of traditional two-way ranging techniques.