PURPOSE: While preoperative prognostic nutritional index (PNI) is a well-established prognostic marker in colorectal cancer (CRC), and postoperative PNI has gained attention, their combined prognostic value remains largely unexplored. PATIENTS AND METHODS: We analyzed patients who underwent curative surgery for stage I-III CRC between March 2004 and February 2014. The pre- and postoperative PNI, measured within 1 month before and 3-8 weeks after surgery, were combined to create "change-PNI" The Cox proportional hazards model was used to assess the prognostic significance, and the C-index was compared across values. RESULTS: The optimal pre- and postoperative PNI cutoff values predicting 5-year overall survival (OS) were 48.05 and 43.65, respectively. The patients were categorized into four groups based on their pre- and postoperative values: pre-low + post-low (G1), pre-low + post-high (G2), pre-high + post-low (G3), and pre-high + post-high (G4). A multivariable Cox proportional hazards model demonstrated that patients in G2, G3, and G4 had significantly lower mortality risks than those in G1 (HR [95% CI] vs G1: G2, 0.341 [0.186-0.625]; G3, 0.457 [0.222-0.941]; G4, 0.222 [0.123-0.401]). The C-index of change-PNI (0.671, 95% CI 0.617-0.720) was superior to that of preoperative PNI (0.609, 95% CI 0.563-0.654) (bootstrap mean difference: 0.062, 95% CI 0.029-0.099) and postoperative PNI (0.622, 95% CI 0.581-0.664) (bootstrap mean difference: 0.049, 95% CI 0.014-0.085). CONCLUSION: Change-PNI serves as a more effective independent immuno-nutritional marker than pre- or postoperative PNI in predicting OS in patients undergoing surgery for non-metastatic colorectal cancer.