The recent success of anti-programmed death 1 (PD1) drugs in treatment of metastatic colorectal cancer patients with mismatch repair deficiency generated overwhelming enthusiasm for immunotherapy in this condition. However, patients with mismatch repair deficient colorectal cancer account for a small proportion of the colorectal cancer population. Current research focuses on advancing immunotherapy to earlier stages of the disease including adjuvant and first-line metastatic settings, and on inducing sensitivity to immune checkpoint inhibitor therapy through a combinatorial approach. However, which patients can benefit from the immunotherapy is an issue needing to be addressed because of the autoimmune toxicity of these drugs. As a detection biomarker, the programmed death ligand 1 (PD-L1) that is one of the ligands for PD-1 can be detected by immunohistochemistry. However, there are some confounding factors in the immunohistochemical detection, such as the application of different detection antibodies, different immunohistochemical threshold values, different collection and preparation methods of tumor tissues, different processing protocols, primary and secondary biopsy specimens, tumor-derived or induced PD-L1 expression, and staining of tumor and immune cells. The current results indicate that patients with tumor overexpression of PD-L1 by immunohistochemistry have better clinical effect when receiving anti-PDL1 treatment, while some tumors with low expression also have remission for this treatment, which lead to the complexities in PD-L1 analysis. Elucidation of the mechanism of host immune system and tumor microenvironment can better explain whether or not the drugs targeting PD-L1 are beneficial for the patients.