Objective To investigate the value of Neutrophil-to-Lymphocyte in predicting the prognosis of pa-tients with severe pneumonia.Methods From January 2021 to July 2023,patients with severe pneumonia were included in Department of Intensive Care Unit in Hangzhou Hospital of Traditional Chinese Medicine.Basic clinical data and bio-markers at day 1,2,3 and 5 were collected to calculate NLR and LMR.According to the outcome of day 28,the patients were divided into survival group and death group.Logistic regression was conducted to analyze independent risk factors,and ROC was conducted to analyze the predictive value of independent risk factors to prognosis.On day 1,2,3 and 5,the AUC of ROC curve for mortality predicted by the NLR was 0.847,0.855,0.879,and 0.954.Results The NLR of the death group patients showed a fluctuating upward trend 5 days before admission,and was higher than that of the survival group;LMR remained consistently lower than the survival group.Binary logistic regression analysis showed that NLR on day 1,2,3,and 5,as well as LMR on day 1,2,and 5,were independent risk factors for all-cause mortality in patients at day 28.NLR had good predictive value for day 28 mortality in patients with severe pneumonia.Conclusion Among pa-tients with severe pneumonia,NLR and LMR are usefur indexes in predecting the death of such patients.