Background: Prediction and potential prevention of sudden cardiac death (SCD) due to malignant ventricular arrhythmia (MVA) represent an obvious unmet medical need. We estimated the prognostic relevance of numerous biomarkers associated with future MVA development in patients with coronary artery disease (CAD) over 2 years of follow-up. Methods: Patients with stable documented CAD (n = 97) with a mean age of 61 ± 10 years were prospectively enrolled in a single-center observational cohort study. Heart failure was diagnosed in 68% of the patients (NYHA class II-III). The mean left ventricular ejection fraction (LVEF) was 50 ± 13%, while 20% of patients had LVEF ≤35%. Sixty-two patients underwent myocardial revascularization during the follow-up (mean 25 ± 11 months). Clinical characteristics (age, gender, diabetes, history of coronary disease and arrhythmias, prior interventions and antecedent medications), noninvasive electrophysiological markers [microvolt T-wave alterations, signal-averaged electrocardiography, QT interval duration and alteration, and heart rate turbulence (HRT) and HR variability], laboratory indices [serum creatinine and creatinine clearance, brain natriuretic peptide (BNP), NT-proBNP, and C-reactive protein and troponin T levels] were assessed with regard to the MVA prognosis. Results: MVA was diagnosed in 11 patients during the prospective follow-up. Prior percutaneous coronary intervention (p < 0.05), MVA or syncope (p < 0.05), on-pump coronary artery bypass grafting during follow-up (p < 0.01), LVEF ≤47% (p < 0.01), a left atrium size ≥4.7 cm (p < 0.05), left atrium index (p = 0.01), filtered QRS duration (p < 0.05), abnormal HRT (χ2 = 6.2, p = 0.01) or turbulence slope (χ2 = 9.5, p < 0.01), BNP ≥158 pg/ml (p < 0.01) and NT-proBNP ≥787 pg/ml (χ2 = 4.4, p < 0.05) were significantly associated with MVA risk by univariate analysis. However, only prior MVA or syncope [odds ratio (OR) 11.1; 95% confidence interval (CI) 2.8-44.4; p < 0.01], abnormal HRT (ОR 13.6; 95% CI 2.8-66.1; p < 0.01) and plasma BNP (ОR 14.3; 95% CI 3.2-65.0; p < 0.01) remained independent predictors of MVA occurrence by multivariate Cox regression analysis. Conclusion: Prior syncope or MVA, HRT and elevated plasma BNP were independent MVA predictors, advocating for the prospective screening of high-risk CAD patients for potential SCD awareness.