Abstract Background and Aims The relation between chronic kidney disease and adverse cardiovascular events is well-established. It is also known that traditional risk factors of cardiovascular disease such as smoking, low-density lipoprotein (LDL) cholesterol and increased blood pressure, cannot fully explain the increased cardiovascular burden observed in patients with chronic kidney disease. In addition, patients with chronic kidney disease do not demonstrate typical clinical symptoms of cardiovascular disease. Thus, good biomarkers for identifying patients at risk and a better understanding of the pathophysiology leading to cardiovascular disease in patients with chronic kidney disease are needed. The objective of the present study was to investigate associations between plasma metabolites and prevalent cardiovascular disease, as well as subclinical cardiovascular disease measured as coronary artery calcification score (CACS), in patients with chronic kidney disease. Method More than 200 metabolic biomarkers, including subclasses of lipoproteins as well as the lipid composition of these, were quantified using nuclear magnetic resonance spectroscopy in 725 patients and 174 controls from the Copenhagen Chronic Kidney Disease Cohort, a single-centre prospective, observational study of non-dialysis patients with stage 1-5 chronic kidney disease. Associations between metabolites and prevalent cardiovascular disease and between metabolites and CACS were determined using multivariable logistic regression and linear regression, respectively. The statistical models were adjusted for traditional cardiovascular risk factors and multiple testing. CACS was determined by CT-scannning and calcium scores were subsequently divided into categories of 0 (no calcification), 1-100, 101-400 and > 400. Results When comparing metabolite concentrations in patients with controls, patients presented with the expected pattern of dyslipidaemia in CKD. We found that they had increased plasma triglyceride concentrations, mainly due to an increase in the triglyceride concentration in very low-density lipoprotein (VLDL) particles, while the concentration of cholesterol in high-density lipoprotein (HDL) particles was decreased. Overall, 85 metabolites were significantly associated with prevalent cardiovascular disease in a model adjusted for eGFR, age and sex (p < 0.001). After further adjusting for diabetes, body mass index, smoking and cholesterol-lowering medication, the significance was lost for all but six metabolites (p < 0.001). The consistent inverse associations with metabolites were primarily involved in HDL metabolism (e.g. ApoA-1, HDL-C and HDL-2). This also applied to the concentration of total lipids in large HDL particles, the concentration of phospholipids in large HDL particles and the the ratio of phospholipids to total lipids in very small VLDL particles. Of the 84 metabolites associated with prevalent cardiovascular disease, 71 were also associated with CACS in a similar pattern. Yet, in the model adjusted for all seven cardiovascular risk factors, only plasma glucose levels as well as the triglyceride content of larger LDL particles remained significant. Conclusion In this study we identified metabolites associated with prevalent cardiovascular disease and subclinical cardiovascular disease (CACS) in patients with CKD. For prevalent cardiovascular disease associations were mainly found for HDL associated metabolites, while CACS was associated with an increase in the triglyceride content of LDL particles and glucose. Further work needs to be done to establish whether these associations are merely a consequence of the cardiovascular burden or whether there is a causal relation.