Although fibronectin has been associated with the pathogenesis of atherosclerosis, little is currently known about the relationship between plasma fibronectin and coronary heart disease (CHD). This retrospective study aimed to determine the predictive value of plasma fibronectin for CHD and its severity. A total of 1644 consecutive patients who underwent selective coronary angiography were recruited into the present study. The characteristics and results of the clinical examination of all patients were collected. Logistic regression analyses were performed to determine the predictive value of plasma fibronectin for the presence and severity of CHD. Compared with non-CHD patients, the CHD patients showed significantly higher plasma levels of troponin I and creatine kinase isoenzyme, along with lower plasma levels of fibronectin. However, no significant differences were detected in plasma fibronectin among patients with different grades of CHD. The logistic regression model showed that plasma fibronectin remained an independent predictor of CHD after adjustment with a 1.39-fold increased risk for every 1 SD decrease in plasma fibronectin. Nevertheless, plasma fibronectin could not predict the severity of CHD determined by the number of stenosed vessels and the modified Gensini score. This study demonstrated that lower plasma fibronectin might be an independent predictor of CHD, but it may be of no value in predicting the severity of CHD.