OBJECTIVE: To analyse the epidemiological trends of tuberculosis in the Siberian and Far Eastern federal districts, the areas with the highest disease burden in the Russian Federation. METHODS: We applied principal coordinate analysis to study a total of 68 relevant variables on tuberculosis epidemiology, prevention and control. Data on these variables were collected over 2003-2016 in all 21 regions of the Siberian federal district and Far Eastern federal district (total population: 25.5 million) through the federal and departmental reporting system. We identified the regions with a favourable or unfavourable tuberculosis epidemiological profile and ranked them as low or high priority for specific interventions. FINDINGS: The median number of tuberculosis notifications in the regions was 123.3 per 100 000 population (range: 54.5-265.7) in 2003, decreasing to 82.3 per 100 000 (range: 52.9-178.3) in 2016. We found large variations in the tuberculosis epidemiological profile across different regions. The principal coordinate analysis revealed that three aggregated indicators accounted for 55% of the variation. The first coordinate corresponded to tuberculosis prevalence and case notifications in the regions; the second to the severity of the disease among patients; and the third to the percentage of multidrug-resistant tuberculosis among tuberculosis patients. The regions where intervention was most urgently needed were Chukotka Autonomous Okrug, Jewish Autonomous Oblast and Tyva Republic. CONCLUSION: The variability in tuberculosis epidemiology across regions was likely due to differences in the quality of antituberculosis services. Precision in defining necessary interventions, as determined through the principal coordinate analysis approach, can guide focused tuberculosis control efforts.