PURPOSE Low concordance between plasma-based and tissue-based tests for determining the RAS mutational status have been reported in some but not all patients with limited-extent metastatic colorectal cancer (mCRC). In this study, we investigated the relationship between metastatic site and circulating tumor DNA (ctDNA) detection using ctDNA genotyping, an alternative to tissue genotyping for precision oncology. MATERIALS AND METHODS We investigated the relationship between metastatic site and ctDNA detection using Guardant360, a next-generation sequencing ctDNA assay, in mCRC patients with single-organ metastasis in the SCRUM-Japan GOZILA study (UMIN000029315). RESULTS Of 1,187 patients with mCRC enrolled in GOZILA, 138 were eligible (49 with liver-only, 15 with lymph node-only, 27 with peritoneum-only, and 47 with lung-only metastases). The concordance of RAS/ BRAF status between Guaradant360 and tissue in vitro diagnostic tests was 95.9% in patients with liver-only, 80.0% in lymph node-only, 56.0% in peritoneum-only, and 65.9% in lung-only metastases. ctDNA fraction, as measured by the median maximum variant allelic fraction (max VAF), and median number of detected variants were 23.1% and five in liver-only, 6.0% and five in lymph node-only, 0.4% and three in peritoneum-only, and 0.4% and three in lung-only metastases, respectively (all P < .001, Kruskal-Wallis test). Few patients with liver-only (2.0%) and lymph node-only metastasis (13.3%) had a max VAF < 0.2%, which is required to ensure a detection limit of 95%, but max VAF was more frequently < 0.2% in patients with lung-only (27.7%) or peritoneum-only metastasis (29.6%). CONCLUSION Patients with lung-only and peritoneum-only metastatic disease have significantly lower levels of ctDNA, suggesting decreased clinical sensitivity for subclonal variants. This observation suggests that such patients may benefit from concurrent tissue and plasma testing to provide optimal genotyping for subsequent therapy selection.