Background: Cervical cancer is a global health problem. There is a need to improve standard treatment with radiotherapy (RT) and concurrent cisplatin (CT). Tumors are known to recruit myeloid cells from the bone marrow (BMDCs) via the CXCL12/CXCR4 and other chemokine pathways, which in turn influence tumor vascular function and RT response. The objective of this study was to explore the impact of the CXCL12/CXCR4-BMDC axis on clinical outcome in cervical cancer patients treated with RT, and ways of inhibiting this response. Methods: A total of 258 patients with cervical cancer (cT1b-T3b, N0-1, M0) were treated with RT +- concurrent CT. Tumor hypoxia, interstitial fluid pressure (IFP – a biomarker of abnormal vascular function) and CXCR4 expression were measured prior to treatment. The median follow-up was 5.3 years. Orthotopic cervical xenografts were developed from patient biopsies for evaluation of the chemokine inhibitor Plerixafor. Results: The pretreatment bloodpolymorphonuclear neutrophil (PMN) count was normal (<7.5X10⁹/L) in 83% of cases.In patients treated with RT+CT, high pretreatment PMNs were associated with inferior disease-free survival (DFS) but only in the setting of high tumor IFP. This was independent of tumor stage, size and LN status, and was validated in a separate cohort of patients treated with RT alone. Patients with high PMNs displayed high peri-vascular tumor infiltration by CD11b+ or CD66b+ myeloid cells. There was 5-fold variation in CXCR4 mRNA expression by qRT-PCR, and 58% of cases showed high CXCR4 protein expression by IHC.Patient-derived orthotopic cervical tumors were treated with fractionated,image-guided RT and CT +- Plerixafor (5 mg/kg). Combined treatment with Plerixafor produced substantial tumor growth delay and prolonged survival compared to standard RT+CT alone. There was no difference in acute GI toxicity with the addition of Plerixafor to standard treatment. Conclusions: TheCXCL12/CXCR4 pathway and BMDCs influence RT response and clinical outcome in patients with cervical cancer via a vascular-dependent mechanism. Inhibition of CXCL12 signaling can mitigate this effect and should be considered for translation to phase I/II clinical trials.