As tumor immunology research continues, the clinical trials and application of immunotherapy represented by the immunological checkpoint inhibitors have achieved satisfactory results, but its objective response rates are low, and its efficacy remains uncertain. The effectiveness of the markers in predicting the immunotherapy efficacy has been inconclusive. A number of clinical trials have found that tumor mutation burden (TMB) is positively correlated with antigens recognized by T-cells and immunotherapy efficacy. It can be, therefore, used to predict the efficacy of immunological checkpoint inhibitors, but with limitations. This article reviewed the predictive value of TMB in immunotherapy for non-small-cell lung cancer and its limitations.