The aim of the study is to examine the effects of parental class, own education and social class on mortality by cause of death among young men. The study is based on 1990 census records for 186,408 Finnish men, aged 30-34 in 1990, linked with death records for 1991-95 (1530 deaths). Parental class was associated with all-cause (Index of Relative Dissimilarity (IRD) = 12%) and cause-specific mortality before adjustment for one's own social class and education. Adjustment reduced the IRD by 18-58%, depending on the cause of death, and caused mortality differences to disappear for causes other than cardiovascular diseases and those related to alcohol. The unadjusted variation in all-cause mortality by educational attainment (IRD = 33%) and by own class (IRD = 38%) was greater than by parental class. Adjustment for parental class had only a minor effect on the variation by education and own class. Adjustment for own class reduced the IRD for education in all-cause mortality by 31%, in mortality from disease by 17%, and in mortality from external causes by 33%. For own class the reductions in the IRD obtained by adjustment for education were 36%, 46%, and 33%, respectively. The effect of parental class on the mortality of young men is indirect and mainly mediated through its influence on education and social class. The effect of education on mortality is as strong as that of occupation-based social class. These variables are not interchangeable measures of socioeconomic status; they both should be taken into account in studies on inequalities in mortality.