Importance Poor olfaction may be associated with incident coronary heart disease (CHD) in older adults, but empirical evidence is limited. Objective To investigate the association of olfaction with risk of CHD. Design, Setting, and Participants This population-based, retrospective analysis of secondary data from the Atherosclerosis Risk in Communities Study, an ongoing prospective cohort, included 5142 US older adults who did not have CHD and had olfaction as assessed at visit 5. At-risk participants were followed up from visit 5 (June 2011 to September 2013) to the date of the first outcome of interest, death, last contact, or December 31, 2020, whichever came first. Data analysis was conducted from March 2024 to January 2025. Exposure Olfaction was measured using a 12-item odor identification test and defined as good (score, 11-12), moderate (score, 9-10), and poor (score, 0-8). Main Outcomes and Measures The primary outcome was incident adjudicated CHD events. The discrete-time subdistribution hazard model was used to estimate the absolute risk of CHD across olfactory statuses and adjusted marginal risk ratio and 95% CI while accounting for covariates and competing risk of death. Results Of 5142 older adults (mean [SD] age, 75.4 [5.1] years), 3234 (62.9%) were female, 1230 (23.9%) were Black, and 3912 (76.1%) were White. After 9.6 years of follow-up (median [quartile 1-quartile 3]: 8.4 [7.4-8.9]), 280 incident CHD events (5.4%) were identified. Poor olfaction was associated with a higher risk of CHD risk, although the association was attenuated with extended follow-up. Comparing poor with good olfaction, the adjusted marginal risk ratio of CHD was 2.06 (95% CI, 1.04-4.53) at year 2, 2.02 (95% CI, 1.27-3.29) at year 4, 1.59 (95% CI, 1.13-2.35) at year 6, 1.22 (95% CI, 0.88-1.70) at year 8, and 1.08 (95% CI, 0.78-1.44) at year 9. The time-varying associations were confirmed using the period-specific, cause-specific Cox regression. The findings were robust in subgroup and sensitivity analyses. The association between moderate olfaction and CHD risk was weaker but showed a similar time-varying pattern as that observed for poor olfaction. Conclusion and Relevance The results of this cohort study suggest that, for older adults, poor olfaction as assessed by a single smell identification test is associated with a higher CHD risk.