Background. The effect of individual educational attainment on health has been extensively documented in western countries, whereas empirical evidence of education spillover effects in marital dyads is scarce and inconsistent. Methods. A total of 2764 individuals (or 1382 marital dyads) were surveyed in the Shanghai Healthy City Project 2008. Logistic regression models were used for analysis, and all analyses were stratified by gender. Results. Significant protective associations were observed in univariate models linking general health status to the individual’s own educational attainment and to their partner’s educational level. After controlling for presence of chronic conditions, lifestyle factors, and social support, these associations were attenuated. Conclusions. The authors found a gender difference in the association of spouse’s educational attainment with self-rated health. The influence of education on health may be partly mediated by lifestyle and other factors.