There is considerable debate about the association between individual socio-economic status, community socio-economic status and health. The current study examines individual data from a case-control study of stroke (n = 3489) conducted in Auckland, New Zealand. The study sought to identify whether individual socio-economic status (as measured by income from lifetime occupation) and community socio-economic status (measured in a number of ways) predicts the onset of stroke both independently and after controlling for individual risk factors (e.g., smoking, obesity and hypertension). Logistic regression results show that individual socio-economic status and all of the community socio-economic status measures predict the onset of stroke before controlling for individual risk factors. However, there is a high correlation between the various measures of community socio-economic status. Stepwise regression results suggest that average household income is the measure of community-level socio-economic status with the greatest predictive power. The results suggest that individual income and average household income are significant predictors of onset of stroke both independently and after controlling for behavioural and medical risk factors. Logistic regression analysis of the pathway suggests that individual income is a significant predictor of smoking and obesity, and that community socio-economic status is a significant predictor of heart disease, heavy drinking, diabetes, smoking and obesity.