The modified Glasgow prognostic score (mGPS), based on C-reactive protein and albumin levels, is an inflammation-based prognostic tool used in various cancers. However, related research in breast cancer is limited. The present study evaluated the prognostic value of the preoperative mGPS in predicting overall survival (OS) of patients with breast cancer undergoing surgery. A retrospective cohort study was conducted involving 300 patients with breast cancer with up to 10 years of follow-up. Patients were categorized into three groups based on mGPS scores of 0, 1 and 2, and their clinical and pathological data were collected. Kaplan-Meier survival analysis and Cox proportional hazards models were used to assess survival outcomes and identify risk factors associated with higher mGPS scores. A prognostic nomogram was developed based on multivariate analysis to predict 5- and 10-year OS. Patients with high mGPS scores showed significantly poor survival outcomes. The 5- and 10-year survival rates for mGPS 0, 1 and 2 were 80, 70 and 55%, and 71, 55 and 22%, respectively (P<
0.001). Multivariate Cox analysis identified the mGPS, age, smoking, PAM50 and TNM stage as independent predictors of OS. The nomogram based on the mGPS demonstrated good predictive accuracy (concordance index: 0.81) and calibration. The preoperative mGPS is an independent prognostic factor for OS of patients with breast cancer. It is a simple, cost-effective tool that can aid in risk stratification and guide treatment strategies. Further validation in larger cohorts is recommended.