Objective — to analyze mortality rates from individual causes of circulatory system diseases (CSDs), nervous and mental disorders, the change of which could be affected by approaches to ascertaining the first cause of death (FCD), in Russia and in its regions in 2013 and 2017. Material and methods. The data of the Russian Federal State Statistics Service (RFSSS) on the number of people who died by causes of death were used on the basis of its «Brief Nomenclature of Causes of Death». The analysis included the following ICD-10 codes: I25.0, I67, I70, I95-99; G20, 21, G30; F00-F09, F20-F29, F30-F69, F80-F99. Mortality rates from the above causes were calculated for each region. To compare the 2013 and 2017 data, the Crimean Federal District was excluded from the analysis. Standardized mortality rates (SMRs) were calculated using the specially developed computer program «Calculation and analysis of mortality rates and years of life lost due to premature mortality in the subjects of the Russian Federation». The 1976 European Standard was used to estimate SMRs. Results. There was the most pronounced decrease in SMRs in the rubric of other and unspecified circulatory system diseases (–79.3%) and in that of atherosclerotic cardiovascular disease, as described (–53.4%), the most significant increment from Parkinson’s disease (+267.7%) and Alzheimer’s disease (+251.4%), as well as the causes combined in the RFSSS nomenclature to the rubric of other psychoses (+419.3%). From our point of view, this is partially due to more intense administrative attention to cause-death coding when implementing vascular programs. Some of the causes of death, which were previously defined as CSDs (atherosclerosis, atherosclerotic cardiovascular disease, as described, cerebrovascular diseases, etc.), are probably coded as Alzheimer’s disease, Parkinson’s disease, taking into account a considerable increase in mortality from these causes. Conclusion. The findings indirectly favor the hypothesis for the spillover of deaths from one class of causes to another one, by changing the approaches to formulating a postmortem diagnosis, determining and coding FCD. There is no doubt that it is necessary to improve the statistical accounting of mortality rates and to clarify criteria for FCD and that it is feasible to conduct the comparative clinical and epidemiological studies in the regions (in the groups of regions), which are aimed at clarifying the causes of death from certain diseases (including those from the class of CSD, nervous and mental disorders).