Assessment tools and body-composition measurements are useful in diagnosing malnutrition. Which one is better for lung disease patients is unclear. The objectives of the present study are: to assess relationships between different methods of nutritional measurements in lung diseases patients; to determine which one is better in diagnosing malnutrition for lung disease patients; and to determine whether lung cancer patients can be differentiated from benign lung disease patients using different measurements. A total of 96 newly diagnosed primary lung cancer patients in stage IIIB/IV and 52 benign lung disease patients nutritional status were assessed according to the SGA, the scored PG-SGA, and serum albumin, prealbumin, transferrin, hemoglobin, total lymphocyte count, body mass index (BMI), and weight. A total of 40% of lung cancer patients were severely malnourished, with men or elder having a higher rate of malnutrition. Significantly lower values of weight, BMI, total lymphocyte count, transferrin, prealbumin and serum albumin were found for them. Age, sex, weight, weight half year ago and prealbumin are in the regression equation to predict them. For benign lung disease patients, 21.2% were severely malnourished with significantly lower values of weight and transferrin. Age and prealbumin are in the equation to predict severely malnourished benign lung disease patients. The highest receiver operation characteristic area under the curve was found for the PG-SGA score, BMI and weight. PG-SGA global rating, age and iron-transferring protein are in the equation for predicting disease status. The SGA and PG-SGA are appropriate for identifying malnutrition in lung disease patients. Lung cancer patients can be differentiated from benign conditions by PG-SGA.