Background: The Homeobox A9 (HOXA9) gene may act as a tumor suppressor in carcinomas and induce breast cancer progression. The clinical potential of methylated HOXA9 (meth-HOXA9) as a biomarker is unclear. Therefore, this study aims to examine the association between meth-HOXA9 in tumor tissue, meth-HOXA9 in plasma, and mortality. Additionally, the study examines the association between meth-HOXA9 in metastatic tissue and plasma at the time of recurrence. Methods: The study enrolled 51 patients with recurrent breast cancer between 2011 and 2015 recruited at the Department of Oncology, Vejle Hospital, University Hospital of Southern Denmark. Tissue samples were obtained from the primary tumor at the time of primary surgery between 1993 and 2013. Tissue samples from the metastases and plasma samples were obtained at the time of recurrence. The study follow-up period ran until death or January 2019. All samples were analyzed for meth-HOXA9 using droplet digital PCR. The optimal cut-off for meth-HOXA9 in tissue was established at ≥7.4% using tissue specimens from independent breast cancer patients and healthy controls. In plasma samples, the detection of ≥5 meth-HOXA9-containing droplets resulted in a positive test. Using Cox regression, we estimated crude and age-adjusted hazard ratios (HR) for mortality with 95% confidence intervals (CI) comparing patients with detectable and undetectable meth-HOXA9 in both tumor tissue and plasma. Fischer’s exact test was used to compare plasma and metastasis meth-HOXA9 status. Results: Among the 50 patients with data on tumor tissue meth-HOXA9, 80% had detectable meth-HOXA9 in the primary tumor, but there was no increased mortality when comparing patients with detectable and undetectable meth-HOXA9 (HR 1.09; CI 0.47-2.52). Only 34 patients had data on plasma meth-HOXA9 and 62% of these had detectable meth-HOXA9 at the time of recurrence. Patients with detectable meth-HOXA9 in plasma had higher mortality than those without (HR 3.95; CI 1.50-10.37). Among the 20 patients with data on plasma and metastatic tissue meth-HOXA9, there was no significant association between meth-HOXA9 in plasma and metastatic tissue (p = 1.00). Conclusions: These results indicate that detectable plasma meth-HOXA9 at the time of breast cancer recurrence is associated with increased mortality. However, there is no association between tumor tissue meth-HOXA9 and mortality. Metastatic tissue and plasma meth-HOXA9 are not associated. Our data support the potential relevance of plasma meth-HOXA9 as a biomarker for prognosis and mortality in recurrent breast cancer, which calls for further investigation.