Background: Studies in multiple tumor types have demonstrated the prognostic impact of ctDNA analysis after curative intent surgery. Emerging data suggest that conversion of ctDNA from detectable to undetectable after adjuvant chemotherapy may reflect treatment efficacy. Further review of existing study data could increase the precision of ctDNA guided adjuvant therapy decision making. Methods: We combined individual patient (pt) data from three independent prospective cohort studies that enrolled 485 pts with stage II or III CRC. Clinicians were blinded to ctDNA results. We evaluated pt outcomes over a 5-year follow-up period (median, 47.2 months). Plasma samples were collected 4 to 10 weeks after surgery. Mutations in ctDNA were assayed using Safe-SeqS. Results: ctDNA was detected after surgery in 59 (12%) pts overall (11.0%, 12.5% and 13.8% respectively for samples taken at 4-6, 6-8 and 8-10 weeks; P = 0.740). ctDNA detection was associated with nodal status; 8.7%, 16.7% and 32.4% in N0, N1 and N2 disease (P < 0.001), but remained an independent adverse prognostic factor in multivariable analysis. ctDNA detection was associated with poor overall survival for pts treated (mortality ratio, 3.0; P = 0.026) or not treated with adjuvant chemotherapy (mortality ratio, 5.17; P < 0.001). The median MAF (mutant allele frequency) in pts with detectable ctDNA was 0.046%. For pts not treated with adjuvant chemotherapy, 3 year recurrence free survival (RFS) was 9% in pts with a MAF > 0.046% vs 33% with a MAF ≤ 0.046% (HR, 2.7; P = 0.032). For chemotherapy treated pts, 3 year RFS was 25% in pts with a MAF > 0.046% vs 70% with a MAF ≤ 0.046 (HR, 3.1; P = 0.025). In 90 pts with recurrence, ctDNA had been detected post surgery in 3 of 20 (15%) with locoregional recurrence, 27 of 60 (45%) with distant recurrence and 5 of 10 (50%) with both (P = 0.044). Conclusions: Where samples for ctDNA analysis were collected 4 to 10 weeks post surgery, sampling timing may not significantly impact detection rates. The prognostic significance of ctDNA detection can be further stratified by MAF level, but MAF level may not impact adjuvant treatment benefit. ctDNA analysis is most sensitive for detecting minimal residual disease at distant sites.