There are many electrocardiographic (ECG) changes in non-ST segment elevation myocardial infarction (NSTEMI). However, the diagnostic power is limited in determining the severity of coronary artery disease (CAD) and clinical outcomes. This study investigated the role of a risk-based ECG score in predicting the severity of CAD and clinical outcomes in NSTEMI patients. One hundred and fifty-two patients were enrolled in the study. Severe CAD was defined as; intermediate (> 22) or high SYNTAX score (> 32), three-vessel disease, and left main coronary artery lesions. A risk-based ECG score was calculated, and the patients were categorized. All patients were followed up, and mortality and repeat revascularizations were evaluated. The severe CAD group had a significantly higher risk-based ECG score than the non-severe CAD group (p = 0.013). The patients with a high risk-based ECG score had more severe CAD (p = 0.013), higher SYNTAX score (p < 0.001), more three-vessel disease (p = 0.003), coronary artery calcification (p = 0.02), and one-year mortality (p = 0.006) than those with medium or low ECG scores. Multivariate logistic regression analysis showed that a 1-point increase in the risk-based ECG score was associated with a 1.573-fold [95% confidence interval (CI): 1.111-2.227, p = 0.011] increase probability of severe CAD. Kaplan-Meier analysis demonstrated that the high-risk group had a significantly higher one-year mortality rate than the low-risk and moderate-risk groups (hazard ratio: 2.383, 95% CI: 1.395-4.072, p = 0.001). This study demonstrated that higher ECG scores were associated with a higher risk of severe CAD and worse clinical outcomes in NSTEMI patients.