Abstract Funding Acknowledgements Type of funding sources: Public Institution(s). Main funding source(s): British Heart Foundation. Background Metabolomics is the detailed profiling of an individual’s circulating metabolites including carbohydrates, lipids, amino and fatty acids, and ketone bodies. Of all the -omics technologies, it provides the most proximal molecular fingerprint to clinical phenotypes and, in conjunction with large-scale population scale cohorts, can allow interrogation of alterations in metabolism that may contribute to cardiovascular disease pathogenesis prior to disease onset. Purpose Given the well-established role of lipids as an important cardiovascular risk factor, this study examines the association between non-lipid circulating metabolites and prognostically important CMR phenotypes in order to provide insight into biological mechanisms as well as identify potentially novel biomarkers. Methods This study used data from the UK Biobank; 31 non-lipid metabolites were measured. Outcome variables were LV end-diastolic volume, LV mass and LV ejection fraction. Metabolite biomarkers were log-transformed and were included in multivariate regression models with additional covariates: age, sex, ethnicity, height, BMI, hypertension, diabetes, hypercholesterolaemia, physical activity, smoking status and fasting time. Correction for multiple testing was performed using the Benjamini-Hochberg method. Results After exclusions and outlier removal, there were 10,769 participants included in the analysis. The mean age was 64 years and 52% of participants were female. Significant results are detailed in Table 1. There was evidence for associations between LV mass, end-diastolic volume and ejection fraction and circulating metabolites, in particular amino acids and fatty acids. Conclusions In the context of the limitations and biases inherent in observational analyses, this study suggests that non-lipid circulating metabolites may influence prognostically important CMR phenotypes. Further exploration, in particular through Mendelian randomisation which can suggest potentially causal associations, is required which may, in turn, provide further biological mechanistic insights into remodelling, as well as highlight potentially novel biomarkers.