Objective The aim of this study was to evaluate the association between pretreatment random plasma glucose levels and cancer prognosis in non-diabetic women with locally advanced cervical cancer treated with CCRT (concurrent chemoradiation) or RT (radiation therapy) only. Methods We investigated the non-fasting plasma glucose levels checked during the initial work up before treatment in 134 non-diabetic patients with locally advanced cervical cancer. Based on the survival time and the progression-free interval (PFI) recorded in the electronic medical records Cox proportional hazard regression models were used to estimate the hazard ratio (HR) for overall survival and PFI according to the various level of glucose and a cut-off level (< 102 mg/dL and ≥ 102 mg/dL), adjusting for clinical covariates. Results A shorter overall survival and PFI was observed in the group with higher glucose levels (HR, 1.03; p = 0.002, HR, 1.02; p = 0.001, respectively) and more than 102 mg/dL, by univariate analyses (HR, 3.21; p = 0.012, HR, 2.20; p = 0.006, respectively). Multivariate analysis, adjusting for clinical FIGO stage, performance status, treatment type (CCRT vs. RT) and chemotherapeutic regimen types showed that patients with higher glucose levels or more than ≥ 102 mg/dL had shorter overall survival times (HR, 1.02; p = 0.015, HR, 2.54; p = 0.049, respectively) and PFI (HR, 1.02; p = 0.003, HR, 1.88; p = 0.031, respectively). Conclusion This investigation provides evidence supporting the prognostic value of glucose levels in non-diabetic women with locally advanced cervical cancer treated with radiation therapy and/or concurrent chemotherapy; high glucose levels were associated with a greater risk for recurrence and mortality in these patients.