Accurately tracking the trajectory of launching vehicles during the boost phase is crucial for a variety of reasons. First, an accurate burnout point location and  velocity of the launch vehicle will facilitate the estimation of its trajectory during the following phases of ballistic flight and the impact location. This has important implications for safety purposes. The first stage of many launcher systems such as the Russian Soyuz or the Chinese Long March rockets commonly impact on land, sometimes on populated areas. Furthermore, the proliferation of suborbital rocket launches presents a hazard if regulations are not properly followed. Therefore, it becomes critical to estimate the impact point of any vehicle that reenters the lower layers of the atmosphere beforehand, during the boost phase, in order to prevent personal and material damage. Second, a critical aspect in any space mission is to insert the payload into the required orbit. An accurate tracking of the launcher system during the boost phase allows carrying out early orbit determination in order to successfully complete the payload insertion. In this Master Thesis, an analysis and performance comparison of different tracking algorithms for launching vehicles during boost (e.g. launcher systems or suborbital rockets) has been carried out. This type of tracking problem has several difficulties that must be overcome. First, the unknown thrust profile – thrust magnitude and orientation – of a launching vehicle makes difficult to develop motion models that can accurately describe its behavior. Second, the observations of the plume of the launching vehicle obtained from two line-of-sight passive sensors located in geostationary satellites are used. The way these observations are obtained does not allow to accurately measuring the initial state and trajectory of the vehicle, complicating the initialization and the trajectory estimation processes of the tracking filters. Finally, the nonlinearities present in the system and measurements models of the tracking filters compel us to devise linearization schemes or the implementation of alternative filtering methods.The work presented in this document has been developed at the Military and  Security department of Airbus Defense and Space, as a continuation of an ongoing project. To solve the aforementioned difficulties of the tracking problem, a tracking system – EKF-Tool – was developed at ADS previous to the start of this Thesis. Nevertheless, this tool shows several limitations that hinder its performance, even preventing from successful tracking. In this Thesis, the limitations and potential areas of improvement of the EKF-Tool have been analyzed. Based on the results of this analysis, an alternative tracking algorithm – PKF-Tool – is introduced with the aim of overcoming the limitations of the EKF-Tool. The tracking performance of both tools has been tested using several indicators: stability, consistency, credibility, accuracy, precision, and versatility.