During the last decades ion beam therapy has become an important treatment method in oncology, especially for patients with deep seated and radio-resistant tumors. One established way of treatment is conventional radiation therapy, using photons or electrons. However, due to its favorable depth-dose distribution, healthy tissue can be spared better when using protons or light ions. While electrons have a short range and X-rays deposit most of their energy shortly after the surface, charged particles have a dose distribution with a peak - the Bragg Peak - at a certain depth, which depends on the projectile, its kinetic energy and the target properties. Additionally, charged particles transfer little energy to the tissue on their passage through it, so they spare organs outside the targeted region. In carbon-beam therapy, ions are used instead of protons, due to much less scattering and the higher biological impact at the terminal depth. However, when heavy ions collide with atoms, nuclear reactions can produce lighter ions that might travel further than the primary terminal depth. This contributes to a dose beyond the Bragg Peak and causes a so called fragmentation tail. Production of secondary particles is highly dependent on the total reaction cross section, which in turn, is dependent on the projectile type and energy and the target. In order to calculate the applied dose in treatment planning, knowledge of the differential and double differential cross sections is necessary. Unfortunately, there are only little experimental data available currently. Until more data can be deduced from measurements it is possible to use Monte Carlo simulation software to estimate the desired cross sections. There are several software packages available, that can simulate particle transport and nuclear reactions, like Geant4 , FLUKA , PHITS , MCNP or SHIELD-HIT . The core of this work was to produce differential and double differential cross section estimates from Geant4 simulations of a carbon particle beam passing through a thin target. These simulations had to be repeated for several iterations of projectile energies, as well as target materials. To obtain the cross sections, an analysis program was developed, that can filter the simulation output for specific particles, based on their type, energy and location. After counting the filtered particles, the program can calculate both types of cross sections and store them on the hard disk. Results of the comparison between simulated data and data from literature will be presented. Additionally, the influence of parameters such as target material , the angle of detection and the ejected particle type on the agreement between data and simulation will be discussed.