Objective: The clinical relevance of the serrated pathway of colorectal carcinogenesis is evident but the screening of serrated lesions remains challenging. We aimed to characterize the serum methylome of the serrated pathway, and to evaluate circulating cell-free DNA (cfDNA) methylation as a source of biomarkers for the non-invasive screening and diagnosis of serrated lesions. Design: We collected serum samples from individuals with serrated adenocarcinoma (SAC), traditional serrated adenomas, sessile serrated lesions, hyperplastic polyps and with no colorectal findings. First, epigenome-wide methylation was quantified in cfDNA pooled samples with the MethylationEPIC array. Then, methylation profiles were compared to tissue and serum cfDNA datasets. Finally, biomarker utility of serum cfDNA methylation was evaluated. Results: We identified a differential methylation profile that can distinguish high-risk serrated lesions from the absence of serrated neoplasia, showing concordance with tissue methylation from SAC and sessile serrated lesions (external datasets). We report that the methylation profiles in serum cfDNA are pathway-specific, clearly separating serrated lesions from conventional adenomas. Among the differentially methylated regions (DMR) we report, the combination of two DMRs within the genes NINJ2 and ERICH1 discriminated high-risk serrated lesions and SAC with 91.4% sensitivity and 64.4% specificity, while methylation from a DMR within ZNF718 reported 100% sensitivity for the detection of SAC at 96% specificity. Conclusion: This is the only study available to date exploring the serum methylome of serrated lesions. We have identified a differential methylation profile in serum specific to the serrated pathway. The serum methylome may serve as a source of non-invasive biomarkers for screening and detection of high-risk serrated lesions.