Precise real-time GPS orbit at a high rate is required for a number of applications, including real-time Precise Point Positioning (PPP), long range RTK and weather forecasts. To support these applications, the International GNSS Service (IGS) has developed a precise orbital service. At present, users may take advantage of the predicted part of the IGS ultra-rapid orbit for real-time and near real-time applications. Unfortunately, however the data rate of such precise orbits is usually limited to 15 minutes. In addition, the precision of the predicted part of the IGS ultrarapid orbit is limited to about 10 cm. for the 24-hour predicted part, which may not be sufficient for the above applications, This research proposes algorithms for interpolation and prediction methods that are intended to reduce the effect of such limitations.This research examines the performance of four interpolation methods for IGS precise GPS orbits, nameley Lagrange, Newton Divided Difference, Bernese Polynomial, Cubic Spline and Trigonometric Interpolation. In addition, a comparison between this research and earlier studies were conducted. A new approach that utilizes the residuals between the broadcast and precise ephemeris to generate a high-density precise ephemeris is also introduced in this research.A three-step neural network-based model is then developed in this research to generate a 6-hour predicted orbital arc. First, an initial predicted orbit is generated by extrapolating a concentrated group of previous precise ephemeris for 5 days. GPS observations for 35 globally distributed tracking stations, corresponding to the 24-hour period preceding the predicted part, are then utilized within the Bernese software to further enhance the predicted orbit. FInally, the predicted orbit is refined by implementing a modular - three-layer feed-forward back-propagation neural network. A comparison is made between our predicted orbit and the IGS ultra-rapid orbit to verify the efficiency of the newly developed neural network-based model. It is shown that the newly developed neural network-based model improved the orbit prediction by 47%, 22% and 37% for three randomly selected satellites from Blocks IIA, IIR and IIR-M respectively.