Objective Explore the relationship between skeletal muscle index at the third lumbar vertebra L3 -SMI and white ball ratio and advanced non-small cell lung cancer NSCLC malnutrition to clarify the efficacy of L3 -SMI and white ball ratio to predict severe malnutrition status in patients with advanced NSCLC. Method Data of 217 advanced NSCLC patients collected in Taiyuan Central Hospital from February 2021 to February 2024 were divided into two groups according to PG - SGA score namely severe malnutrition group n = 43 grade C and non-severe malnutrition n = 174 grade A+B comparing the demographics and clinical characteristics of the two groups univariate and binary Logistic regression analysis to affect independent factors of patients ' malnutrition using nomogram analysis model L3 -SMI and A/ G ratio by ROC curve analysis. Result Compared with non -severe malnutrition later clinical stage higher age and lower A/ G ratio L3 -SMI in severe malnutrition P<0. 05 . Further binary Logistic regression analysis indicated that clinical stage Ⅳ older age lower A/ G ratio and lower L3 -SMI were independent factors affecting severe malnutrition in advanced NSCLC patients. The nomogram was constructed from independent factors with C- index of 0. 99 model threshold probability range from 1% to 99% with net yield> 0 absolute agreement between true and predicted values of 0. 011 and prediction model AUC of 0. 942 95% CI 0. 879 - 0. 999. Meanwhile by the ROC curve analysis the A/ G ratio and L3 - SMI predicted the AUC of severe malnutrition in advanced NSCLC were 0. 919 and 0. 908 respectively. Conclusion L3 -SMI and A/ G ratio is the main evaluation index of severe malnutrition in patients with advanced NSCLC which has high prediction efficacy. At the same time clinical stage of advanced and late are also independent factors that induce severe malnutrition in patients. By focusing on monitoring the change of L3 - SMI and A/ G ratio levels it can detect body abnormalities in time to intervene and improve the prognosis.