Peripheral blood-derived inflammation-based markers, including C-reactive protein (CRP), neutrophil-to-lymphocyte ratio (NLR), lymphocyte-to-monocyte ratio (LMR), and platelet-to-lymphocyte ratio (PLR) are indicators of prognosis in various malignant tumors. The present study aimed to identify the inflammation-based parameters that are most suitable for predicting outcomes in patients with breast cancer. Two hundred ninety-six patients who underwent surgery for localized breast cancer were reviewed retrospectively. The association between clinicopathological factors and inflammation-based parameters were investigated. Univariate and multivariate Cox regression analyses were performed to identify independent prognostic indicators associated with disease-free survival (DFS). The NLR level correlated significantly with tumor size (P<0.05). The PLR level correlated with the expression of estrogen receptor and lymph node involvement (P<0.05). Univariate analysis revealed that lower CRP and PLR values as well as tumor size, lymph node involvement, and nuclear grade were significantly associated with superior DFS (CRP: P<0.01; PLR, tumor size, lymph node involvement, and nuclear grade: P<0.05). On multivariate analysis, CRP (hazard ratio [HR]: 2.85, 95% confidence interval [CI]: 1.03-7.88, P<0.05), PLR (HR: 2.61, 95% CI: 1.07-6.36, P<0.05) and nuclear grade (HR: 3.066, 95% CI: 1.26-7.49, P<0.05) were significant prognostic indicators of DFS in patients with breast cancer. Neither LMR nor NLR significantly predicted DFS. Both preoperative CRP and PLR values were independently associated with poor prognosis in patients with breast carcinoma; these were superior to other inflammation-based scores in terms of prognostic ability.