The risk of developing adverse side effects in the normal tissue after radiotherapy is often limiting for the dose that can be applied to the tumor. Proton minibeam radiotherapy, a spatially fractionated radiotherapy method using sub-millimeter proton beams, similar to grid therapy or microbeam radiation radiotherapy (MRT) using X-rays, has recently been invented at the ion microprobe SNAKE in Munich. The aim of this new concept is to minimize normal tissue injuries in the entrance channel and especially in the skin by irradiating only a small percentage of the cells in the total irradiation field, while maintaining tumor control via a homogeneous dose in the tumor, just like in conventional broad beam radiotherapy. This can be achieved by optimizing minibeam sizes and distances according to the prevailing tumor size and depth such that after widening of the minibeams due to proton interactions in the tissue, the overlapping minibeams produce a homogeneous dose distribution throughout the tumor. The aim of this work was to elucidate the prospects of minibeam radiation therapy compared to conventional homogeneous broad beam radiotherapy in theory and in experimental studies at the ion microprobe SNAKE. Treatment plans for model tumors of different sizes and depths were created using the planning software LAP-CERR, to elaborate suitable minibeam sizes and distances for the individual tumors. Radiotherapy-relevant inter-beam distances required to obtain a homogeneous dose in the target volume were found to be in the millimeter range. First experiments using proton minibeams of only 10 µm and 50 µm size (termed microchannels in the corresponding publication Zlobinskaya et al. 2013) and therapy-conform larger dimensions of 100 µm and 180 µm were performed in the artificial human in-vitro skin model EpiDermFT^TM (MatTek). The corresponding inter-beam distances were 500 μm, 1mm and 1.8mm, respectively, leading to irradiation of only a few percent of the cells in the skin tissue, but with significantly increased doses (up to 5000 Gy) compared to the average dose of 2 Gy, which was applied homogeneously in further skin samples for comparison. Gaussian-shaped minibeams of even larger sizes (σ=260 µm and 520 µm, inter-beam distance 1.8mm) were analyzed in further experiments to evaluate the effect of increasing beam sizes as in deeper-lying tissues. Acute side effects were quantified via the MTT tissue viability test and the release of inflammatory proteins into the culture medium and showed improved results for minibeam compared to homogeneous irradiation. Genetic damage, an indicator for secondary tumor induction, was analyzed via the micronucleus test in the epidermal keratinocytes and was less than half for minibeams up to 180 µm size compared to homogeneous fields. Increasing minibeam sizes, i.e. increasing fractions of irradiated skin receiving a dose higher than the average dose of 2 Gy) increased the number of micronuclei per divided cell, but never exceeded the genetic damage induced by a homogeneous dose distribution. A more authentic and representative in-vivo skin model, accounting for higher complexity with blood vessels, further cell types, follicles glands and especially a working immune system, was used in the next step to further examine the side effects of minibeam radiotherapy compared to homogeneous irradiation. The central part of the ear of adult BALB/c mice was irradiated with 20 MeV protons, using an average dose of 60 Gy in a field of 7.2 x 7.2mm². The 4 x 4 minibeams of nominal 6000 Gy had a size of 180x180 µm² and inter-beam distances of 1.8 mm, as in previous in-vitro skin experiments. Minibeam irradiation induced no ear swelling or other visible skin reaction at any time, while significant ear swelling (up to 4-fold), skin reddening (erythema) and desquamation developed in homogeneously irradiated ears 3-4 weeks after irradiation. Loss of hair and sebaceous glands only occurred in the homogeneous irradiation fields and did not recover during the monitoring phase of 90 days. Taken together all theoretical considerations and experimental findings, proton minibeam radiation therapy appears suitable for the implementation in clinical tumor therapy using protons and/or heavy ions, as it reduces side effects in the normal tissue compared to conventional broad beam irradiation. However, the upper limit of the minibeam size for tissue sparing and the technical feasibility are still to be elucidated as current technologies might have to be improved and adapted for the generation of sub-millimeter proton beams of energies up to 250 MeV at therapy plants.