Aims: The aim of this study was to test whether the association between self-rated health and mortality differs between educational groups in Norway, and to examine whether health problems and health-related behaviour can explain any of these differences within a previously unexplored contextual setting. Methods: The study used data from the Nord-Trøndelag Health Study 84-86 (HUNT) with a 20-year follow up. The analyses were performed for respondents between 25–101 years at baseline ( n = 56,788). The association between self-rated health and mortality was tested using Cox regression. Results: The results indicate that although self-rated health is associated with mortality there is no difference in the association between self-rated health and mortality between educational groups. Introducing health-related variables did not have an impact on the result. Conclusions: Given the small educational differences in the association between self-rated health and mortality, this supports the reliability of self-reported health as a measurement for objective health.