Sibling correlations have gained increasing prominence in inequality studies as a measurement of the total impact of family background on individual outcomes. Whilst previous studies have tended to use traditional socio-economic measures such as parent’s income or education, this paper introduces an analytical class approach to sibling studies by analysing how much of the influence that siblings share in their long-term income results from class origin. Data are from Statistics Denmark and consist of 290,399 individuals born between 1963 and 1973. Models are estimated which – in addition to parents’ education and income – include modifications of the Erikson–Goldthorpe–Portocarero schemes ranging from 3 to 15 classes and Grusky’s micro-class scheme of 72 classes. The results show that although class adds to explanations of the family influence on children’s income, most of the sibling similarities are not explained by parental education, income or class. Depending on gender, the class schemes explain between 8 and 13 per cent of the sibling similarities and 15 to 20 per cent when parents’ income and educations are also included. Models with different class schemes demonstrate that elaborated versions of the EGP class scheme add little to the explanation of similarities between brothers, sisters and mixed siblings.