Objectives To investigate the consistency between the Patient-Generated Subjective Global Assessment (PG-SGA) and the Global Leadership Initiative on Malnutrition (GLIM) criteria in preoperative nutritional evaluation of colorectal tumor patients undergoing radical surgery, based on clinical survey data.Methods Data were collected from 105 patients who underwent laparoscopic radical resection of colorectal carcinoma in the Second affiliated Hospital of Wenzhou Medical University between July 2021 and Nov 2022. Preoperative PG-SGA and GLIM assessments were conducted, and the agreement between the two methods was analyzed using the Kappa test. The GLIM results were further adjusted by integrating biochemical indicators (prealbumin, albumin, and creatinine) and handgrip strength measurements to evaluate changes in agreement.Results Without adjustment, the agreement between PG-SGA and GLIM was poor (Kappa=0.171, P<0.01). Following physiological and biochemical parameter adjustments, the agreement remained weak (Kappa=0.382), but improved significantly to strong agreement after secondary calibration with handgrip strength (Kappa=0.771, P<0.001). Significant differences in nutritional assessment outcomes were observed between age groups (≥70 years vs. <70 years, P< 0.05), while no statistically significant differences were found between different gender groups (P>0.05).Conclusion Both PG-SGA and GLIM are effective tools for preoperative nutritional assessment, though discrepancies exist in malnutrition classification criteria. However, their complementary nature allows for significantly improved agreement through a multidimensional adjustment approach "physiological and biochemical corrections + muscle function enhancement". It is recommended to integrate both systems by utilizing GLIM for standardized malnutrition diagnosis and PG-SGA for its dynamic scoring features, thereby providing evidence-based support for personalized nutritional intervention strategies and postoperative clinical outcome optimization. Future efforts should focus on expanding sample sizes and developing dynamic monitoring models to facilitate updates in clinical guidelines.