For the time being most external radiation therapy treatments employ photons. Over the last decades particle beam therapy became more widely available, utilizing accelerated protons and carbon ions in cancer management. The clinical experience gained with protons and carbon ions stimulates the refinement of treatment schemes and the evaluation of treatment related side effects. It seems clear that no particle alone can be optimal for all clinical situations. Therefore, in the last years, a thorough investigation of novel ions and on the optimal ion selection has begun. In this PhD thesis the feasibility of particle beam therapy employing helium ions is investigated. In a first step, the physical properties of helium ions were benchmarked against protons employing Monte Carlo simulations. Range straggling effects and lateral beam broadening were reduced by a factor of two. A flexible pencil beam (PB) algorithm was developed. It was based on the idea of fluence weighted elemental PB kernels. A novel minimization based splitting approach allowed the calculation of arbitrary beam shapes. Look-up tables (LUTs) were used to determine the dose deposition alongside the penetration path. Beam broadening due to multiple Coulomb scattering was accounted for and additional correction terms were included using a Voigt function. Verification showed excellent agreement in homogeneous phantoms and heterogeneous phantoms consisting of layered slabs. Dose calculation precision of the developed PB algorithm was similar for proton beams and helium ions. Its design provides easy customization and the use of non-Gaussian beam profiles, while allowing competitive calculation speeds. Currently available data on relative biological effectiveness (RBE) of helium ions alongside with linear energy transfer (LET) evaluations motivated an empirical zonal RBE model. A RBE of 1.0 in the plateau region, was followed by a rise up to 2.8. For protons, increasing evidence also points towards a dynamically changing RBE. Therefore, the same model concept with a maximum RBE of 1.6 was used. The resulting dose calculation module was integrated into the treatment planning system Hyperion. A prospective treatment planning study was conducted. Representative patients of different treatment sites, i.e. prostate, base-of-skull, paediatric, and head-and-neck tumours were included. Biologically optimized plans as well as plans employing only physical dose contributions were created and evaluated. For both particles planning target volume parameters were similar. Using helium ions doses to organs at risk (OAR) could be reduced. Biological optimized treatment plans showed bigger reductions in OAR dose compared to plans with only physical dose contributions. The data outlined the potential benefits of helium ions, but did not allow quantitative evaluations. ^Therefore, a planning study with a larger patient collective was performed for protons and helium ions. Three different paediatric indications, Neuroblastoma (NB), Hodgkin lymphoma and Wilms tumour were studied. Differences were highest for NB patients, where the body volume receiving less than 60% of the prescribed dose could be reduced by up to 10% employing helium ions. Especially for liver and kidneys pronounced differences were found. Generally, the use of helium ions resulted in improved OAR sparing for selected indications, but reductions were smaller than initially expected. Increased differences could be expected for higher doses and deeper located tumours. In summary, the results indicate that helium ions may provide an attractive alternative to protons in the low LET region. However, other ion species may also be of benefit for radiation therapy. Further careful evaluations of novel ion species are necessary in order to refine biological modelling and balance benefits and potential down-sides such as fragmentation effects or dosimetric challenges.