Importance: Primary resistance to immune checkpoint inhibitors is observed in 10% to 40% of patients with metastatic colorectal cancer (mCRC) displaying microsatellite instability (MSI) or defective mismatch repair (dMMR). Objective: To investigate possible mechanisms underlying primary resistance to immune checkpoint inhibitors of mCRC displaying MSI or dMMR. Design, Setting, and Participants: This post hoc analysis of a single-center, prospective cohort included 38 patients with mCRC diagnosed as MSI or dMMR by local laboratories and entered into trials of immune checkpoint inhibitors between January 1, 2015, and December 31, 2016. The accuracy of MSI or dMMR status was also assessed in a retrospective cohort comprising 93 cases of mCRC that were diagnosed as MSI or dMMR between January 1, 1998, and December 31, 2016, in 6 French hospitals. Primary resistance of mCRC was defined as progressive disease according to Response Evaluation Criteria in Solid Tumors criteria, 6 to 8 weeks after initiation of immune checkpoint inhibitors, without pseudo-progression. All tumor samples were reassessed for dMMR status using immunohistochemistry with antibodies directed against MLH1, MSH2, MSH6, and PMS2, and for MSI using polymerase chain reaction with pentaplex markers and with the HSP110 T17 (HT17) repeat. Main Outcomes and Measures: The primary outcome was positive predictive value. Results: Among the 38 patients (15 women and 23 men; mean [SD] age, 55.6 [13.7] years) in the study with mCRC displaying MSI or dMMR, primary resistance to immune checkpoint inhibitors was observed in 5 individuals (13%). Reassessment of the status of MSI or dMMR revealed that 3 (60%) of these 5 resistant tumors were microsatellite stable or displayed proficient mismatch repair. The positive predictive value of MSI or dMMR status assessed by local laboratories was therefore 92.1% (95% CI, 78.5%-98.0%). In the retrospective cohort of 93 patients (44 women and 49 men; mean [SD] age, 56.8 [18.3] years) without immune checkpoint inhibitor treatment, misdiagnosis of the MSI or dMMR status by local assessment was 10% (n = 9), with a positive predictive value of 90.3% (95% CI, 82.4%-95.0%). Testing for MSI with the HT17 assay confirmed the MSI or dMMR status in 2 of 4 cases showing discrepant results between immunohistochemistry and pentaplex polymerase chain reaction (ie, dMMR but microsatellite stable). Conclusions and Relevance: Primary resistance of mCRC displaying MSI or dMMR to immune checkpoint inhibitors is due mainly to misdiagnosis of their MSI or dMMR status. Larger studies are required to confirm these findings. Microsatellite instability or dMMR status should be tested routinely using both immunohistochemistry and polymerase chain reaction methods prior to treatment with immune checkpoint inhibitors.