Objective To explore the predictive value of neutrophil-to-lymphocyte ratio on the prognosis of H7N9 avian influenza. Methods A retrospective analysis was conducted on 28 H7N9 avian influenza patients (treatment group) at the First Affiliated Hospital of Soochow University from April 2013 to January 2016. Thirty healthy physical examiners in the same period were enrolled as the healthy control group. The 28 patients were followed up for half a year and divided into the improvement group (18 cases) and the death group (10 cases) according to the clinical prognosis. Inflammatory indicators including white blood cells (WBC), neutrophil (N), lymphocyte (L), monocytes (M), platelet (PLT), creatine kinase (CK), lactate dehydrogenase (LDH), high sensitive C reactive protein were collected at day 1, day 3 and week 1 of admission. Calculation of neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR), ΔNLR3 (day 3 of admission NLR-on day 1 of admission NLR), ΔNLR7 (week 1 of admission NLR -day 3 of admission NLR) and so on calculating ΔPLR3, ΔPLR7, ΔLMR3, ΔLMR7. Differences of the above indicators between the improvement group and death group were compared. The measurement data with normal distribution were tested by t-test of two independent samples, and the count data with non-normal distribution were tested by Mann-Whitney U-test. Univariate and multivariate logistic regression analysis to explore the prognostic factors and the working characteristic curve of subjects was used to evaluate the predictive value of inflammatory response indexes for H7N9 avian influenza death. Results In the treatment group, the baseline WBC, L, N, PLT, the proportion of lymphocytes, neutrophils, monocytes, and NLR, PLR, and LMR were all statistically different compared with the healthy control group (all P<0.01). After treatment, day 3 NLR, ΔNLR3 in improvement group were both significantly decreased to 10.93 (15.71) and 0.87 (-15.63), respectively when compared with death group (17.62[23.63] and 7.42[22.68], respectively) (Z=-2.16 and -2.014, respectively, both P<0.05). Day 7 NLR, ΔNLR7 in improved group were 6.51 (13.23) and -0.37 (-12.38), respectively, which were both lower than those of death group (27.90 [25.64] and 11.54 [-26.22]) with statistically significant differences (Z=-2.444 and -2.111, respectively, both P<0.05). Multivariate logistic regression analysis indicates that ΔNLR3 is the main factor that affects the prognosis of the H7N9 infection (odds ratio [OR]=1.153, 95% confidence interval [CI]: 1.052-1.263, P=0.002). Reciver operating characteristic curve analysis showed that the area under the curve was 0.733 (95% CI: 0.532-0.935, P=0.044). Based on the principle of Youden index, the cut-off value of ΔNLR3 to predict the death risk of H7N9 avian influenza was 5.453 with sensitivity of 0.700 and the specificity of 0.722. The mortality was higher when ΔNLR3 was higher than 5.453. Conclusions Dynamic monitoring NLR, especially ΔNLR3 may reflect the condition and prognosis of H7N9 infection, which is an independent predictor of death. Key words: H7N9 avian influenza virus; Neutrophil lymphocyte ratio; Prognosis