Objective To study the effect of serum non-high-density lipoprotein cholesterol (non-HDL-C) on the occurrence of coronary heart disease (CHD) and the severity of coronary artery stenosis in elderly non-smokers. Methods A total of 360 elderly non-smokers (≥65 years old) who admitted to Chengde Central Hospital due to chest pain and underwent coronary angiography from June 2016 to June 2019 were recruited in this study. According to having CHD or not, they were divided into non-CHD group (n=110) and CHD group (n=250). The 250 cases of CHD patients were further assigned into mild lesion group and severe lesion group according to the severity of coronary artery stenosis (Gensini score), and into single, double, multi vessel lesion group based on the number of involved coronary arterial branches. The clinical data, biochemical indicators, and results of coronary angiography were compared among the subgroups. SPSS statistics 25.0 was used for data analysis. Student′s t test, ANOVA, Mann-Whitney U test or Chi-square test was employed for inter-group comparison. Spearman correlation analysis was adopted to analyze the correlation between non-HDL-C and Gensini score, logistic regression analysis was used to analyze the risk factors of severe CHD, and receiver operating characteristic (ROC) curve was drawn to evaluate the diagnostic efficiency of CHD. Results The levels of uric Acid (UA), lipoprotein-a (Lp-a), triglyceride (TG), total cholesterol (TC), HDL-C, and non-HDL-C were significantly higher in the CHD group than the non-CHD group (P<0.05). Compared with the mild lesion group, the severe lesion group had higher levels of Lp-a, TC, HDL-C and non-HDL-C (P<0.05). With more branches involved in coronary artery disease, non-HDL-C level and Gensini score were significantly increased; and with the increase of serum non-HDL-C level, the Gensini score was also elevated (P<0.05). Spearman correlation analysis showed that non-HDL-C level was significantly positively correlated with Gensini score (r=0.812, P＜0.001). Logistic regression analysis indicated that non-HDL-C was an independent risk factor for severe coronary artery disease in CHD patients. The ROC curve suggested that when the serum non-HDL-C level was 3.58 mmol/L, the diagnostic efficiency was the highest, with an area under the curve of 0.837 (95%CI 0.803-0.870), a sensitivity of 64.20% and a specificity of 96.70%. Conclusion Non-smoking elderly CHD patients are more common with multivessel coronary artery lesions and severe stenosis. Serum non-HDL-C is an independent risk factor affecting the severity of coronary artery disease in them, and shows a high efficiency in the diagnosis of CHD.