When measuring social differences in mortality in the former socialist countries of central and eastern Europe, it has been impossible, until now, to link census data and vital records to avoid the classic problem of bias caused by inconsistency between the individual status recorded in the census and that reported at the time of death. The results presented here come from one of the first studies ever conducted in a former Soviet-bloc country based on linkage of individual vital records with census data, in this case deaths in 2001-2004 and the 2001 census. The study considers cause-specific mortality differentials among persons over age 30 by educational level, marital status, ethnicity and place of residence. The results differ from those that would have been obtained with the traditional aggregative approach but confirm the existence of sharp inequalities, notably for infectious diseases and the effects of alcoholism.