Abstract BACKGROUND AND AIMS Patients with late-stage chronic kidney disease (CKD) who are non–dialysis-dependent (NDD) or incident to dialysis (ID) (e.g. initiated dialysis within the last 4 months) or dialysis-dependent (DD) represent a vulnerable population at increased risk for morbidity and mortality despite significant clinical advancements in recent years. The safety of roxadustat in patients with NDD or ID-DD CKD requires further elucidation. METHOD In this post-hoc exploratory analysis, safety results from eligible patients with anemia of CKD enrolled in four phase 3, randomized, open-label studies [NDD (DOLOMITES) or ID-DD (SIERRAS, HIMALAYAS, ROCKIES)] were pooled and compared roxadustat to an erythropoiesis-stimulating agent (ESA). Endpoints were time to major adverse cardiovascular event [MACE; myocardial infarction, stroke and all-cause mortality (ACM)] and MACE+ (MACE plus congestive heart failure or unstable angina requiring hospitalization), ACM, and treatment-emergent adverse events (TEAEs). MACE and MACE + were evaluated for non-inferiority using hazard ratios (HRs) at 1.8 and 1.3-upper margins, respectively, and 95% confidence intervals (CIs). TEAEs were descriptively summarized. RESULTS In total, 2142 patients were evaluated (1083 roxadustat; 1059 ESA; 616 NDD; 1526 ID-DD). Roxadustat was non-inferior to ESA for risk of MACE [HR 0.79 (95% CI 0.61–1.02)] and MACE+ [HR 0.78, (95% CI 0.62–0.98)] with a consistent finding for ACM [HR 0.78 (95% CI 0.57–1.05)]. TEAEs were generally comparable between roxadustat and ESA groups, including any TEAE [incidence rate per 100 patient-exposure years (IR/100 PEY) 56.1 versus 53.5], TEAEs leading to discontinuation of study drug (IR/100 PEY 6.7 versus 5.1) and TEAEs leading to death (IR/100 PEY 6.9 versus 7.4). The most frequent (IR/100 PEY) TEAEs were hypertension (roxadustat 12.8, ESA 12.3), end-stage kidney disease (roxadustat 6.6, ESA 6.1), diarrhea (roxadustat 7.1, ESA 4.8) and hyperkalemia (roxadustat 4.3, ESA 4.8). CONCLUSION There was no evidence of an increased risk of cardiovascular events or mortality with roxadustat compared with ESA in patients with anemia who have NDD or ID-DD CKD. Although TEAE development occurred commonly in both the roxadustat and ESA groups, patients infrequently discontinued a study drug because of an adverse event.