In this paper, an alternative batch filter has been presented and utilized for precise satellite orbit determination. The alternative batch filter is developed without any linearization and approximation process using the Unscented Transformation. For the precise orbit determination system, the position, velocity and a parameter for ballistic coefficient of the low earth orbiting satellite at epoch time are considered as states to be estimate. The range, azimuth, and elevation angles of satellite measured from three ground tracking stations are used for observation. For evaluation and verification of the presented batch filter performance. The results are compared with those of a batch least squares filter for various initial errors in position, velocity, and measurement sampling periods. For relatively small initial errors in position and velocity or short measurement sampling periods, the accuracy of orbit determination is similar in both the filters. Under large initial errors in position and velocity or long measurement sampling periods, the alternative batch filter yields more robust and stable convergence than the batch least squares filter. Furthermore, the alternative batch filter can be easily utilized because it does not require calculations of partial derivatives. The alternative batch filter based on the Unscented Transformation is effectively applicable for highly nonlinear batch estimation problems.