Background <p>Emerging evidence suggests metabolic dysregulation may contribute to colorectal cancer (CRC) aetiology. We aimed to identify pre-diagnostic metabolic biomarkers for CRC risk in 230,420 UK Biobank participants.</p> Methods <p>Nuclear magnetic resonance spectroscopy was used to quantify 249 metabolic biomarkers in plasma samples collected at baseline. Cox proportional hazards models were used to estimate hazard ratios and 95% confidence intervals (CIs) for associations of metabolic biomarkers with CRC risk after adjusting for potential confounders. To infer the potential causality of biomarkers that were associated with CRC independent of the others, we performed genome-wide association analyses among 199,732 UK Biobank participants of European ancestry to identify biomarker-associated genetic variants, followed by two-sample Mendelian randomization (MR) analyses using summary statistics of 78,473 CRC cases and 107,143 controls of European ancestry.</p> Results <p>During a median follow-up time of 9.7 years, 2,410 incident primary CRC cases were identified. Among 43 CRC-associated (<i>P</i>-value &lt; 0.001) metabolic biomarkers, ten biomarkers including fatty acids (FAs), inflammation, ketone bodies, and lipoprotein lipids were associated with CRC risk after mutual adjustment. MR analyses provided strong evidence for potential causal associations of CRC risk with percentages of linolic acid [odds ratio (OR) = 0.89, 95% CI = 0.83-0.96, <i>P</i>-value = 3 × 10^-3] and saturated FAs (OR = 1.14, 95% CI = 1.03–1.25, <i>P</i>-value = 9  ×  10^-3) to total FAs.</p> Conclusions <p>We identified multiple CRC-associated metabolic biomarkers. Perturbed lipid and lipoprotein metabolism may promote colorectal carcinogenesis.</p> <p></p>