New challenging dosimetric approaches, such as narrow beams and 3D algorithms, are being used in radiotherapy. In this paper two quality control (QC) procedures are reported. The first one concerns the QC of the dosimetry of small x-ray beams, generally carried out by using silicon detectors. The comparison of dose values obtained by a silicon diode, a diamond detector, and radiochromic films shows that for x-ray beams of high energy, the silicon diode can give an overestimation of the output factors in phantom, up to 4%. This is due to the higher than unit density silicon diode and the surrounding envelope that restore the lateral electron equilibrium. About the 3D algorithms for breast treatment planning, a quality control test has been adopted to verify the accuracy of the computed dosimetry when "loss of scatter" occurs. The results show a sensible agreement (within 1.5%) between computed and experimental data.