Background Having regained its political autonomy in 1991, Estonia experienced major changes in political, economic, and social realities. We aimed to analyse mortality changes by education from 1989 to 2000 in order to assess the impact of recent changes in Estonia, as well as the delayed effects of pre-transitional developments. Methods Two census-based analyses were compared. Individual cause-specific death data for those aged 20+ for 1987–1990 (72 003 deaths) and 1999–2000 (35 477 deaths) came from the national mortality database. Population denominators came from the population censuses of 1989 and 2000. Mortality for all causes combined and for selected causes of death were analysed for high, mid, and low educational groups. The absolute differences in mortality were evaluated through life expectancy at age 25 and age-standardized mortality rates. To assess the relative differences between educational levels, mortality rate ratios with 95% CI were calculated using Poisson regression. Results Educational differences in mortality increased tremendously from 1989 to 2000: over the 10-year period life expectancy improved considerably for graduates, and worsened for those with the lowest education. In 2000, male graduates aged 25 could expect to live 13.1 years longer than corresponding men with the lowest education; among women the difference was 8.6 years. Large differences were observed in all selected causes of death in 1989 and in 2000 and the trends were invariably much more favourable for the higher educated. Educational differences in total mortality increased in all age groups. Conclusions Social disruption and increasing inequalities in wealth can be considered main recent determinants; however, causal processes, shaped decades before recent reforms, also contribute to this widening gap.