Objective To analyze the correlation between body nutritional risk and prognosis in elderly patients with severe pneumonia, and to study the application value of geriatric nutritional risk index (GNRI) for elderly patients with severe pneumonia. Methods A total of 133 elderly patients with severe pneumonia admitted to Foshan Fosun Chancheng Hospital from January 2020 to January 2022 were recruited, and according to their GNRI within 24 h after admission, the patients with high nutritional risk were assigned into high-risk group, and the other patients were into other classification groups. The patients with high nutritional risk were given nutritional support treatment, and based on the intervention timing of nutritional support, they were divided into early-, middle- and late-stage subgroups. The prognosis was compared among the patients of different timing of nutritional support. After the 28-day mortality rate was counted, the patients were grouped into death and survival groups. SPSS statistics 19.0 was used to process the data. Data comparison between two groups was perfomed using t test orχ² test depending on data type.Binary logistic regression model was applied to analyze the related factors affecting the prognosis (death within 28 d) in elderly patients with severe pneumonia, and receiver operating characteristic (ROC) curve was drawn to analyze the value of GNRI in predicting the 28-day mortality in these elderly patients. Results Among the 133 elderly patients with severe pneumonia, there were 28 cases (21.05%) with high nutritional risk. The patients with high nutritional risk had advanced age, higher detection rate of Gram-negative bacteria, larger ratio of mechanical ventilation and higher incidence of shock, but lower body mass index, serum albumin and pre-albumin levels when compared with the other classification groups (P<0.05). After the patients with high nutritional risk were treated with nutritional support, the times for mechanical ventilation and vasoactive drug use and length of hospital stay were all in increasing trends in the patients with early, middle and late nutritional interventions in turn, and the death rate was lower in the patients with early intervention than those with late intervention (P<0.05). Among the 133 elderly patients with severe pneumonia, 35 cases (26.32%) eventually died within 28 d after admission. Binary logistic regression analysis confirmed that Acute Physiology and Chronic Health Evaluation Ⅱ (APACHE Ⅱ) score at admission (OR=1.906, 95%CI 1.350-2.691), multiple organ dysfunction syndrome (MODS) score (OR=2.079, 95%CI 1.254-3.448), mechanical ventilation (OR=2.177, 95%CI 1.313-3.610) and high-risk GNRI grade (OR=2.575, 95%CI 1.778-3.730) were risk factors for 28-day death in elderly patients with severe pneumonia. ROC curve analysis found that the AUC value of GNRI (AUC=0.706, 95%CI 0.607-0.806) in predicting 28-day death in elderly patients with severe pneumonia was significantly higher (P<0.05) than that of APACHE Ⅱ score (AUC=0.534, 95%CI 0.428-0.641), MODS score (AUC=0.564, 95%CI 0.460-0.668) and mechanical ventilation (AUC=0.628, 95%CI 0.518-0.737). Conclusion It is suggested that GNRI should be used for nutritional risk screening in elderly patients with severe pneumonia as soon as possible after admission, and nutritional intervention should be carried out in time for those with high-risk malnutrition so as to improve the prognosis of patients.