BACKGROUND. Life expectancy at birth reflects to some extent the effectiveness of public health system. In Ukraine, life expectancy is not equal in different regions: the level varies from 66.5 years in the Donetsk region to 71.5 years in Kiev. Therefore, the purpose of the study was to determine the indicators of public health system functioning that explain such interregional differentiation.METHODS. Empirical data analyzed in the study included annual regional statistics provided by the State Statistics Committee of Ukraine for 24 Ukrainian regions, the Autonomous Republic of Crimea, and cities Kyiv and Sevastopol in 2001-2008. The model includes the following variables: life expectancy at birth (dependent variable), the number of places in sanatoria, the amount of local spending on health in the current and the previous year, the number of hospital beds, the number of doctors of all specialties, the number of ambulance stations, average duration of hospitalization, average wages, and the ecological situation in the region. The relationship was assessed using multivariate linear regression analysis.RESULTS. The results show that regions with higher life expectancy at birth are better equipped with doctors and ambulance stations, but have lower average duration of hospitalization and lower average wages.CONCLUSIONS. Not all the revealed associations may be interpreted as causal. While we can hypothesize that doctors and ambulances may contribute to higher life expectancy at birth, still there may exist unmeasured characteristics of regions which make them attractive for health workers. Higher salaries are observed in Eastern regions of Ukraine; however, poorer health there may be caused not by higher salaries but by higher prevalence of unhealthy behaviors and environmental problems. Longer duration of in-patient treatment may be an indicator of inefficacious health care management, on the one hand, but still larger numbers of chronic patients would result in both longer duration of in-patient treatment and shorter life expectancy, on the other hand. The conducted study gives grounds for certain hypotheses which can be tested with other data.