Differences in mortality by socio-economic position (SEP) are well established, but there is uncertainty as to which dimension of SEP is most important in what context. This study compares the relationship between three SEP dimensions and mortality in Finland, during the periods 1990–97 and 2000–07, and to existing results for Sweden. We use an 11% random sample from the Finnish population with information on education, occupational class, individual income and mortality (age groups 35–59 and 60–84) (n = 810,902; 274,316 deaths). Cox proportional hazard models produce hazard ratios (HR) for categories of SEP variables in bivariate and multivariate models. Multivariate HRs are smaller than bivariate HRs, but all dimensions have a net effect on mortality. Overall, income shows the steepest mortality gradient: HR = 2.49 among men in the lowest income quintile aged 35–59 in the 1990s. The importance of the various SEP dimensions is modified by gender and age group, reflecting the significance of gendered life course differences in analyses of health inequality. Except for the declining disadvantage of poor men aged 35–59, inequalities are very stable over time and similar between Finland and Sweden. In such studies, the use of only one SEP indicator functions well as a broad marker of SEP. However, only analyses of multiple dimensions allow for comprehensive measurements of SEP, take into account the fact that some SEP dimensions are mediated by others, and provide insights into the social mechanisms underlying the stable structure of inequalities in mortality.