Based on the findings from the Radiation Effects Research Foundation's studies of the cohort of Japanese atomic bomb survivors, it has been reported that total-body irradiation at 0.5–1.0 Gy could be responsible for increased rates of mortality from broad-based categories of cardiovascular disease (CVD), i.e., stroke and heart disease. However, CVD consists of various subtypes that have potentially different radiation dose responses, as well as subtype-specific risks that have not been fully evaluated. Potential problems with changes in the coding rules for the International Classification of Diseases (ICD) and the underlying causes and trends in CVD mortality in Japan also need to be considered. The goal of this study was to clarify the radiation risk of subtype-specific heart disease over different time periods. Radiation dose response was examined for mortality from several heart disease subtypes in 86,600 members of the Life Span Study (LSS) cohort during 1950–2008. These subtypes included ischemic heart disease (IHD), valvular heart disease (VHD), hypertensive organ damage (HOD) and heart failure (HF). Individual radiation doses ranged between 0 and 4 Gy. In addition to analyses for the total period, we examined specific periods, 1950–1968, 1969–1980, 1981–1994 and 1995–2008, corresponding to major developments in medical technologies and ICD code revisions. We observed significant positive associations between radiation dose and mortality from heart disease overall in 1950–2008 [excess relative risk or ERR/Gy (95% CI) = 0.14 (0.06, 0.22)]. Subtype-specific ERRs also positively increased with dose: 0.45 (0.13, 0.85) for VHD, 0.36 (0.10, 0.68) for HOD and 0.21 (0.07, 0.37) for HF, respectively. No significant departure from linearity was shown for the dose-response model. Although there was no evidence for a threshold in a model function, the lowest dose ranges with a statistically significant dose response were 0–0.7 Gy for heart disease overall and VHD, 0–1.5 Gy for HOD and 0–0.4 Gy for HF. No significant association between radiation exposure and IHD was observed in any model, although a quadratic model fit the best. The risk of HOD and rheumatic VHD increased significantly in the earliest periods [ERR/Gy = 0.59 (0.07, 1.32) and 1.34 (0.24, 3.16), respectively]. The risk of nonrheumatic VHD increased with calendar time and was significant in the latest period [ERR/Gy = 0.75 (0.02, 1.92)]. The risk of IHD, especially for myocardial infarction, tended to be elevated in the most recent period after 2001, where cautious interpretation is needed due to the uncertain validity of death diagnosis. Radiation risks of heart disease mortality in the LSS appeared to vary substantially among subtypes, indicating possible differences in radiation-induced pathogenesis. Trends in CVD rates in Japan during the long observation period may also impact risk analyses.