We investigated red blood cell (RBC) PUFA profiles, and the predictive value of RBC EPA content for tumour EPA exposure and clinical outcomes, in the EMT study, a randomised trial of EPA in patients awaiting colorectal cancer (CRC) liver metastasis surgery (Cockbain et al., 2014) [8]. There was a significant increase in RBC EPA in the EPA group (n=43; median intervention 30 days; mean absolute 1.26[±0.14]% increase; P<0.001), but not in the placebo arm (n=45). EPA incorporation varied widely in EPA users and was not explained by treatment duration or compliance. There was little evidence of 'contamination' in the placebo group. The EPA level predicted tumour EPA content (r=0.36; P=0.03). Participants with post-treatment EPA≥1.22% (n=49) had improved OS compared with EPA <1.22% (n=29; HR 0.42[95%CI 0.16-0.95]). RBC EPA content should be evaluated as a biomarker of tumour exposure and clinical outcomes in future EPA trials in CRC patients.