Objective. To assess executive function in healthy adults using fMRI. Material and methods. An analysis of fMRI activation and functional connectivity during a serial count task (as a shifting function test) and color-word Stroop test (classical inhibition function test) was made for 12 healthy adults. Results and conclusion. The executive control network and salience network activation was comparable in both tasks. Nevertheless, there were differences between two tests in functional connectivity of the dorsolateral prefrontal cortex (DLPFC) and the supplementary motor area (SMA) with other brain regions, that can be explained by the differences in the regulatory mechanisms of task performance. Stroop test assumes its automatic performance, and control of program realization is performed mainly by executive-control network. The connectivity between the two DLPFCs with the lower parietal lobules and with each other and inhibition by SMA connectivity with only the right hemisphere regions support this notion. Serial count task excludes the process of monotonous learning, that was confirmed by widespread SMA connections in the absence of connectivity of the DLPFC with executive control network regions. This connectivity pattern allows assuming the leading role of SMA in certain brain regions choice and switching their activity for providing attention and executive control of cognitive operations shift during task performance. These findings allow us to consider the serial count task as the relevant fMRI test for executive functions with the special focus on set shifting, also in patients with executive function deficits. Furthermore, SMA region mapping with the serial count test paradigm could be considered as a potential target for navigated transcranial magnetic stimulation (nTMS) in these patients.