[en] Intensity-modulated radiotherapy (IMRT) is an advanced form of 3-D conformal radiotherapy, that uses a non uniform spatial modifications in the intensity of the beams across the irradiated field. On the basis of studies comparing modulated radiation therapy approach and other treatment modalities, IMRT may improve the dose distributions for target volumes with complex shapes and/or concave regions. Because of the additional degrees of freedom to define the treatment plan respect to conventional radiotherapy techniques, is necessary to use a computerized iterative optimization. In this work we have analysed the results obtained on some cases of head-and-neck tumours comparing the results with data obtained in a previous work on some prostate tumour cases in order to evaluate the intrinsic uncertainty of the optimization process for an IMRT plan. The close proximity of organs at risk (e.g. parotids, sub-mandibular glands for head and neck tumour cases) and the concave shape of the tumour volume make this type of tumour cases well suited for a treatment with IMRT. Because of such complex anatomy, IMRT planning for head and neck tumours is particularly time consuming and challenging. So, it is particularly important to evaluate the optimization process in order to obtain the most efficient treatment plan. Our approach consists of defining a dose-volume cost function, used as quality index of the treatment plan, and minimizing, by means of a downhill iterative procedure, the cost function to achieve the optimal solution. Local minima distribution has been used as an indicator of the complexity of the optimization process. The statistical analyses of the data have shown that the process in IMRT has an intrinsic uncertainty. This uncertainty is represented by the distribution of the possible solutions for the optimization problem in the search space around global and local minima. Different initial conditions and different stopping criteria of the minimization algorithms has been studied to determine the robustness of the approach. (author)