Two competing hypotheses underpinned an investigation into limiting long-standing illness (LLI) among UK graduates. Hypothesis 1 proposed childhood social class (CSC) influences LLI independently of educational attainment and adult income. Hypothesis 2 proposed typical middle-class graduates would have lower LLI prevalence than typical and atypical working-class and atypical middle-class graduates. Working/middleclass refers to childhood circumstances. Atypical/typical refers to full-time employment duration before attending university. Graduates (1985; N = 5093 and 1990; N = 8147) were followed up in 1996. Logistic regression was used to examine LLI in 1996 by CSC only and CSC, atypical graduate status and their interaction, adjusting for age and adult income. Hypothesis 1 was not confirmed. Hypothesis 2 was partially confirmed. Typical middle-class graduates had a lower LLI prevalence than typical working-class and atypical middle-class graduates. These results support the idea that opportunities for good human functioning are culturally determined and affect health.