Diabetes mellitus, and even prediabetes, has been shown to be independently associated with cardiovascular disease. Our study explored whether the combination of glycosylated hemoglobin (HbA1c) and fasting blood glucose (FBG) can better assess the severity of coronary heart disease (CHD) in elective percutaneous coronary intervention (PCI) patients. We consecutively enrolled 1006 prediabetic patients with HbA1c 5.7% to 6.4% who underwent elective PCI. Patients were divided into 2 groups: a normal fasting glucose (NFG) group (FBG <6.1 mmol/L) and an impaired fasting glucose (IFG) group (6.1 ≤FBG<7.0 mmol/L) with defined values. Baseline characteristics and angiography data of the 2 groups were compared. The prevalence of 3-vessel disease (P = .002), the GENSINI (the score is named after a professor) score (P = .002), and the SYNTAX (SYNergy between PCI with TAXUS™ and Cardiac Surgery) score (P = .002) of the IFG group was significantly higher compared to the NFG group. After multiple regression analysis, FBG was found to be independently associated with prevalence of 3-vessel disease (adjusted odds ratio: 1.62; 95% confidence interval: 1.21-2.36; P = .013), the GENSINI score (standardized β = .138, P = .008), and the SYNTAX score (standardized β = .145, P = .005). In addition, HbA1c was independently associated with the prevalence of 3-vessel disease, the GENSINI, score, and the SYNTAX score (P < .05). Both FBG and HbA1c are independently correlated with the severity of CHD in prediabetic patients with HbA1c 5.7% to 6.4%.