Introduction For targeted radiation therapies, treatment planning is a real challenge. In fact, the activity that can be injected to the patient is limited by the dose to organs at risk (OARs) which has to be kept below the tolerance threshold. Therefore, a 3D dosimetric method (PMCD), based on Monte Carlo calculations, has been developed in order to personalize and optimize the activity prescription for Selective Internal Radiotherapy (SIRT) treatments. Methods The PMCD method was used with data from 10 patients treated for hepatic metastases. Using the OEDIPE software, regions of interest outlines, drawn on CT images, were used to create patient-specific voxel phantoms and 99mTc-MAA SPECT data were used to generate 3D-matrices of cumulated activity. The absorbed dose and biological effective dose (BED) were calculated at the voxel scale using the MCNPX Monte Carlo transport code and OEDIPE. The maximum injectable activity (MIA) was then calculated using OEDIPE depending on tolerance criteria on OARs based on mean absorbed doses, mean BED, Dose-Volume Histograms (DVHs) or BED-Volume Histograms (BVHs). Those MIAs were compared to the ones recommended by conventional methods: the Body Surface Area (BSA) method and the Partition Model with tolerance criteria on mean absorbed doses. Results Performing a personalized dosimetry using the PMCD method enables to increase the activity prescription while ensuring OARs’ radiation protection. Moreover, considering tolerance criteria based on DVHs further enhances treatment planning efficiency by taking advantage of the parallel characteristic of the liver and the lungs whose functions are not impaired if the level of irradiation to a fraction of the organ is kept below the tolerance threshold. Finally, tolerance criteria on mean BED and BVHs further improve treatment planning by taking into account biological considerations such as cell repair, radiosensibility or dose-rate distribution. Conclusion Besides its feasibility and applicability in clinical routine, the interest of a personalized Monte Carlo dosimetry for treatment planning in 90Y-microspheres therapies was confirmed from those patient studies.