Personalization of selecting a target for rTMS is a problem, solving which can significantly increase the method efficacy. Stimulation of the key hubs of individual-level resting-state networks represents an approach to personalization. The study aimed to develop an rTMS personalization method based on the selection of individual frontoparietal control network (FPCN) hubs and assessment of their localization variability. To determine the FPCN hubs, individual maps were built using the FPCN group mask as a seed. The searchlight algorithm with the sphere radius of 5 mm was used to select targets within the dorsolateral prefrontal cortex (DLPFC) and posterior parietal cortex (PPC). The target spatial localization variability and correctness of using the routine “5 cm rule” for the DLPFC localization were analyzed. In 24 healthy volunteers (9 males, average age 29±7 years), high interindividual variability of targets was demonstrated. In no area is there a universal position of the stimulation coil that would effectively stimulate targets in all volunteers. Spatial dispersion of points is higher in the DLPFC (volumes of the polyhedral containing the point sets are 2095 mm³ in the DLPFC and 739 mm³ in the PPC). All individual targets in the DLPFC are located within the FPCN mask, while in the PPC some targets are outside this mask. The average distance between the М1 zones and DLPFC is 64±13 mm. In 75% of the subjects, this exceeds 5 cm, which confirms that it was incorrect to use the routine “5 cm rule” for coil positioning in the majority of subjects. An algorithm to select personalized targets for rTMS based on the resting-state fMRI data in the DLPFC and PPC being the key FPCN hubs has been developed.