739 Background: Gastrointestinal (GI) cancers account for over one-third of cancer-related deaths. We have previously developed GutSeer, a non-invasive assay utilizing cell-free DNA (cfDNA) methylation and fragmentation signatures, to cost-effectively detect and localize five major GI cancers, including colorectal (CC), gastric (GC), liver (LC), esophageal (EC), and pancreatic cancer (PC). Methods: GutSeer was initially developed and validated in a large retrospective cohort of 1844 plasma samples, demonstrating excellent performance with specificity, sensitivity, and tissue of origin (TOO) accuracy at 96.7%, 86.2%, and 82%, respectively. In this study, we conducted a blind, multi-center clinical test for further validation. Participants were recruited from two centers, Beijing (BJ) and Shanghai (SH). Cohort BJ included 42 participants (20 healthy, 22 cancer), and Cohort SH comprised 403 participants (105 healthy, 123 benign, 175 cancer). Over half of the cancer samples in Cohort SH were in early-stage disease (TNM stages I and II). Results: In Cohort BJ, GutSeer showed 100% specificity, 68.2% sensitivity, and 86.7% TOO accuracy. Individual cancer type sensitivities within Cohort BJ were 100% for CC, 57.1% for EC, 75% for GC, 40% for PC, and 100% for LC. In Cohort SH, GutSeer exhibited 94.6% specificity for healthy samples and 86.2% for benign conditions. For cancer samples in Cohort SH, the overall sensitivity achieved 78.9%, with specific sensitivities of 92.3% for CC, 81.0% for EC, 78.4% for GC, 52.9% for PC, and 86.4% for LC. Moreover, GutSeer also efficiently detected early-stage cancers, with stage-wise sensitivities ranging from 64.7% (stage I) to 96.3% (stage IV). TOO accuracy in Cohort SH was 79.0%, with individual cancer types achieving accuracies from 71.7% to 94.7% (72.2% for CC, 71.7% for EC/GC, 94.7% for LC, and 77.8% for PC). Conclusions: GutSeer demonstrates promising potential as a cost-effective, non-invasive test for early detection and localization of major GI cancers. Its consistently high performance in both retrospective and multi-center blind clinical tests implies its potential to revolutionize the screening and diagnosis paradigm for GI cancers.