Coronary artery disease (CAD), the leading cause of morbidity and mortality, has imposed huge health and economic burdens globally. Zinc-α2-glycoprotein (ZAG) is a novel adipokine. Increasing evidence suggests the close relationship between serum ZAG levels and various cardiometabolic risk factors. However, the relationship between serum ZAG levels and CAD is still not fully clarified. We conducted this study to evaluate serum ZAG levels and its association with cardiovascular risk factors. A total of 129 patients with CAD, 99 patients with noncoronary artery disease (NCAD), and 121 controls were recruited in this retrospective study. CAD (coronary artery stenosis ≥50%) or NCAD (coronary artery stenosis <50%) patients who underwent coronary angiography were diagnosed according to the American Heart Association criteria. Serum ZAG levels were determined via commercial enzyme-linked immunosorbent assay (ELISA) kits. e results showed that serum ZAG levels in CAD and NCAD groups were significantly decreased when compared with those in the control group. Multiple stepwise regression analysis revealed that the grouping variable (control, NCAD, and CAD) was an independent determinant of serum ZAG levels (β � − 0.328, P < 0.001) after controlling other confounding factors. Further multivariate ordinary logistic regression analysis demonstrated that the risk of grouping at one level higher in subjects with the lowest tertile of ZAG levels was 2.28-fold higher than those with the highest tertile levels (OR � 3.281, 95% CI 1.782-6.038, P < 0.001). e receiver-operating characteristic (ROC) curve analysis showed that serum ZAG could distinguish CAD patients (AUC � 0.706, 95% CI, 0.643-0.770, P < 0.05), NCAD patients (AUC � 0.673, 95% CI, 0.602-0.743, P < 0.05), and NCAD and CAD patients (AUC � 0.692, 95% CI, 0.633-0.750, P < 0.05) from controls. In conclusion, serum ZAG levels were significantly decreased in NCAD/CAD patients. e decreased serum ZAG levels were independently associated with the presence of NCAD/CAD. ZAG might serve as a candidate diagnostic biomarker for NCAD/CAD.