Aim The effects of family caregiving, especially long‐hours caregiving, on coronary heart disease (CHD) are debatable. We examined the impact of family caregiving on incident non‐fatal CHD. Methods We used data from the Longitudinal Survey of Middle‐Aged and Elderly Persons from 2005 to 2010, a nationwide panel survey for Japanese people aged 50–59 years in 2005 (baseline). After we excluded non‐respondents and people with missing key variables at baseline, 25 121 individuals without CHD, stroke or cancer were followed up for a mean of 4.6 years. The exposure was assessed at baseline by three indicators: (i) family caregiving; (ii) hours spent caregiving; and (iii) kinship type of care recipient. The non‐fatal CHD incidence was identified according to questionnaire responses from 2006 to 2010. Results Cox's proportional hazards analysis did not show a statistically significant association between family caregiving and incident non‐fatal CHD (hazard ratio [HR] 1.13, 95% confidence interval [CI] 0.92–1.40). Caregivers who spent 20–69 h per week on care showed a statistically significant increased risk for non‐fatal CHD (HR 1.78, 95% CI 1.23–2.58) compared with non‐caregivers; whereas this increased risk was statistically significant only among women (HR 1.98, 95% CI 1.27–3.08), but not among men (HR 1.35, 95% CI 0.67–2.71). Kinship type of care recipient did not make a significant difference to the effects of family caregiving on incident non‐fatal CHD. Conclusions Long‐hours family caregiving could be an independent risk factor for incident non‐fatal CHD among middle‐aged women in Japan. Geriatr Gerontol Int 2017; ••: ••–••.