Introduction: In-hospital mortality after isolated coronary artery bypass grafting is now low, yet it differs markedly between centers and countries. Preoperative risk estimates guide patient routing, informed consent, and quality benchmarking. Because surgical technique, perioperative care, and patient profiles have shifted since 2018, risk models built on earlier cohorts lose calibration and call for re-examination in current practice. Objective: To assess contemporary external validity and calibration of EuroSCORE II and the Society of Thoracic Surgeons risk model (STS) for shortterm mortality in adults undergoing isolated coronary artery bypass grafting. Methods: We performed a systematic search of PubMed/MEDLINE, Scopus,and Web of Science restricted to 2018–2025 and English-language publications. Eligible records were original external validations in isolated coronary artery bypass grafting or studies with a clearly reported subgroup; endpoints were inhospital mortality or 30-day survival. We extracted discrimination (area underthe ROC curve) and calibration metrics (observed-to-expected mortality ratio,Hosmer – Lemeshow p-value, calibration slope and intercept when available). Study selection and data extraction were performed independently by two reviewers; unreported metrics were recorded as “not reported”. Risk of bias was assessed using Prediction Model Risk of Bias Assessment Tool (PROBAST). Results: Nine studies from Europe, Asia, and Latin America were included(sample sizes 438–14,118). Discrimination ranged from modest to high: for EuroSCORE II AUC 0.55–0.90 (0.82–0.90 in the largest cohorts); for the STS model AUC 0.65–0.88. Calibration was heterogeneous: overestimation in low-risk and younger patients (e.g., overall observed-to-expected ratio 0.58; 0.26 in patients <60 years) contrasted with underestimation in older adults, in procedures with cardiopulmonary bypass, and in several non-European populations. In some reports, simple centre-specific recalibration improved agreement between predicted and observed outcomes. Conclusion: In contemporary isolated coronary artery bypass grafting, EuroSCORE II and the STS model remain useful for risk ranking but centrelevel calibration checks and – when indicated – straightforward recalibration are recommended. Regular auditing of observed-to-expected mortality and stratified verification in key subgroups is also recommended. Received 22 October 2025. Revised 20 March 2026. Accepted 10 April 2026. FundingThe study did not have sponsorship. Conflict of interestThe authors declare no conflict of interest. Contribution of the authorsConception and study design: L.A. Arutyunyan Data collection and analysis: L.A. Arutyunyan, S.V. Shalaev, S.B. TsiryatievaStatistical analysis: L.A. Arutyunyan, L.V. Kremneva, V.V. Ostanina Drafting the article: L.A. Arutyunyan, R.T. Gaibova, M.V. Martynenko Critical revision of the article: L.A. Arutyunyan, S.V. Shalaev, S.B. Tsiryatieva Final approval of the version to be published: all authors