In Japan, the effects of socioeconomic indicators on coronary risk factors and subjective well-being in an urban population have not been compared. The subjects of this study were 1361 civil servants (968 men and 393 women, aged 35-64 years) working in an urban area of Japan. Screening examinations were conducted from April 1997 to March 1998, and a questionnaire survey was conducted in February 1998. The effects of two socioeconomic indicators, education level (junior high school, high school and university education) and employment grade (manual, low-level nonmanual and high-level nonmanual work), on behavioral and biological coronary risk factors, self-rated health and affect balance were investigated using multivariate logistic regression analyses. Smoking was found to be inversely associated with level of education; compared to university graduates, the odds ratios for subjects who had only graduated from high school and those for subjects who had only graduated from junior high school were 1.96 and 2.07 in men and 3.44 and 5.48 in women, respectively. As for alcohol drinking and physical inactivity, relationships were inconsistent in terms of direction by the two indicators. Among biological risk factors, diabetes was inversely associated with education level in men. Self-rated health, however, was inversely associated with employment grade both in men and women, and affect balance was inversely associated with employment grade in men. In summary, different effects of two socioeconomic indicators, education and employment grade, were seen in some coronary risk factors and subjective well-being in an urban Japanese population. Our findings should contribute to the elucidation of mechanisms of the socioeconomic gradients of risk factors and mortality from coronary heart disease in Japan.