Background: Sleep quality and exercise frequency are closely associated with coronary heart disease (CHD). Few studies focused on the joint effect of initiating sleep, sleep disorders and exercise frequency on the risk of CHD in elderly. Based on Boshan Elderly Cohort Study (BSECS), we explored the sleep quality, exercise frequency and their joint effects on the risk of CHD.Methods: We collected 678 participants whose age ≥60 years old from Boshan district hospital. We used Pittsburgh Sleep Quality Index (PSQI) to evaluate the sleep quality, and obtained physical examination information from hospital.Results: Compared with non-CHD group, CHD patients spent more time in initiating sleep (time≥60mins, 34.59% vs 22.93%, P=0.025), and less time in exercising(exercise frequency<1times/week, 23.90% vs 17.15%, P=0.024). In multiple logistic regression analysis, sleep latency ≥ 60 mins were associated with CHD risk (adjusted OR=1.82; 95%CI: 1.16, 2.86; P=0.010). The adjusted OR (95%CI) of CHD was 2.24 (1.16, 4.34) for sleep duration <5 hours vs 5 - 9 hours. Compared with exercise frequency <1 times/week, the adjusted OR for exercise frequency ≥ 1 times/week was 0.66 (95% CI: 0.44, 0.98; P=0.037). In addition, the joint effects of long sleep latency (≥ 60 mins) and sleep disorders were associated with CHD (adjusted OR=3.29; 95%CI: 1.35,7.98; P=0.009), and the joint effect of exercise frequency ≥ 1 times/week and good initiating sleep were also associated with CHD, the adjusted OR (95%CI) was 0.39 (0.20, 0.77).Conclusions: Long sleep latency, high frequency of initiating sleep difficulty, sleep disorders and short sleep duration were positively associated with CHD. In addition, the joint effect of long sleep latency and sleep disorders were positively correlated with CHD incident, however, the joint effect of exercise frequency ≥ 1times/week and good initiating sleep were negatively associated with the risk of CHD.