Background Recent studies from the USA and Europe suggest an association between an individual's political ideology and their health status, with those claiming to be conservatives reporting better health. The presence of this association is examined in Japan.

Methods Individual-level data from the 2000–3, 2005 and 2006 Japan General Social Survey were analysed. The outcomes of interest were self-rated poor health and smoking status. The independent variable of interest was reported political beliefs on a 5-point ‘left’-to-‘right’ scale. Covariates included age, sex, education, income, occupational status and fixed effects for survey periods. Logistic regression models were estimated.

Results There was an inverse association between political ideology (left to right) and self-rated poor health as well as between ideology and smoking status even after adjusting for age, sex, socioeconomic status and fixed effects for survey periods. Compared with those who identified as ‘left’, the OR for reporting poor health and smoking among those who identified as ‘right’ was 0.86 (95% CI 0.74 to 0.99) and 0.80 (95% CI 0.70 to 0.91), respectively.

Conclusions Health differences by political ideology have typically been interpreted as reflecting socioeconomic differences. The results from Japan corroborate the previous findings from the USA and Europe that socioeconomic differences do not account for health differences by political ideologies. Political ideology is likely to be a marker of several latent values and attitudes (eg, religiosity, individual responsibility and/or community participation) that might be beneficial for health at the individual level.