In modern, advanced radiotherapy (e.g. intensity modulated photon radiotherapy, IMXT) the delivery time for each fraction becomes prolonged to 10-20 minutes compared with the conventional, commonly 2-5 minutes. The biological effect of this prolongation is not fully known. The large number of beam directions in IMXT commonly leads to a large integral dose in the patient. Electrons would reduce the integral dose but are not suitable for treating deep-seated tumour, due to their limited penetration in tissues. By combining electron and photon beams, the dose distributions may be improved compared with either used alone. One obstacle for using electron beams in clinical routine is that there is no available treatment planning systems that optimise electron beam treatments in a similar way as for IMXT. Protons have an even more pronounced dose fall-off, larger penetration depth and less penumbra widening than electrons and are therefore more suitable for advanced radiotherapy. However, proton facilities optimised for advanced radiotherapy are not commonly available. In some instances electron beams may be an acceptable surrogate. The first part of this study is an experimental in vitro study where the situation in a tumour during fractionated radiotherapy is simulated. The effect of the prolonged fraction time is compared with the predictions by radiobiological models. The second part is a treatment planning study to analyse the mixing of electron and photon beams for at complex target volume in comparison with IMXT. In the next step a research version of an electron beam optimiser was used for the improvement of treatment plans. The aim was to develop a method for translating crude energy and intensity matrices for optimised electrons into a deliverable treatment plan without destroying the dose distribution. In the final part, different methods of treating the spinal canal in medulloblastoma were explored in a treatment planning study that was evaluated with biological models for estimating risks for late radiation effects. The effect on cell survival of prolonging fraction time at conventional doses/fraction is significant in an in vitro system. This effect is underestimated by biological models. Prolonging the fraction time will spare tissues with a fast DNA repair. Thus, there is a risk for sparing tumours. The mixed electron and photon beam technique has the potential to treat deep-seated tumours. Compared with IMXT the number of beams can be reduced and as a consequence, the time for each fraction could be kept shorter. The integral dose in the patient will also be lower. The mixed beam technique could potentially be further improved if automatic optimisation for electrons was available. The results suggest that optimisation and segmentation can be automated, and a deliverable treatment plan can be obtained with simple procedures without destroying the quality of the dose distribution. The integral dose in patients may lead to late radiation side effects. In childhood cancers the risk for development of radiation induced cancers is a reality and the integral dose outside the target volume should be minimised. Based on models for cancer induction, protons show the lowest risk while electrons have some benefit compared with different photon techniques. All methods are able to similarly well treat the target volume.