This paper analyses the factors involved in differences in the experience of long-term illness (severe and non-severe illness), as measured in terms of self-reported frequency and intensity of symptoms. The study has a cross-sectional design. It uses a database from the Survey of Living Conditions of Statistics Sweden, and treats a representative sample of the employed Swedish population (n = 13,501), aged between 16 and 65, interviewed over the period 1986-89. The results show that male manual workers report more non-severe and severe illness than non-manual workers, and that manual and lower-level non-manual female workers report more severe illness, but not non-severe illness, than intermediate/higher-level non-manual working females. The observed class differences in experience of severity of illness are partly explained by the factors investigated (job demands, personal economic difficulties, smoking daily, weak social network) in the case of men, and virtually entirely in the case of women. Other ill-health dimensions, such as self-rated general health and impaired working capacity, prove to be related to severity of illness, the latter being more strongly associated with experience of severe illness than the former irrespective of social class. The results lend support to the hypothesis that manual classes are subjected to what might be called "double suffering"; they have more long-term illnesses and also experience these illnesses with greater intensity and frequency.