Choroid plexus tumors (CPTs) are rare neoplasms of the central nervous system most commonly found in the pediatric population. CPTs represent 1- 4% of all childhood brain tumors, with 10- 20% occurring during the first year of life. Within this family of tumors, choroid plexus carcinoma (CPC) is the malignant neoplasm which is categorized as a grade III tumor by the WHO. Choroid plexus papilloma (CPP) is a benign form classified as a grade I tumor, and atypical choroid plexus papilloma (aCPP) as a grade II tumor. Distinction between these tumor subtypes is essential for treatment stratification. Previous studies performed in our laboratory suggest that CPTs are highly unstable and harbor unique patterns of chromosome-wide gains and losses. To better understand the complexities of tumor biology of CPTs as well as to identify better molecular biomarkers to distinguish between aggressive and benign forms of CPTs we performed a genome-wide DNA methylation study using Illumina Human Methylation450 BeadChip. We analyzed genome-wide DNA methylation profiles from 34 CPT (14 CPCs, 5 aCPPs and 15 CPPs) samples. Differential DNA methylation analysis did not identify significant differences between aCPPs and CPPs, therefore we explored CPC-specific DNA methylation signature in comparison to CPPs. Using a median beta value difference of 0.3 or greater and an FDR adjusted p-value<0.05, we identified 3361 CpGs that showed significant difference in methylation between CPCs and CPPs or aCPPs. Two-way clustering performed using Pearsons correlation and average linkage for both the sample tree and the gene tree revealed segregation between the majority of CPCs and CPPs or aCPPs. Two main clusters were discovered within CPCs that were due to differences in TP53 mutation status. Pathway analysis on a 1328 gene set overlapping the 3361 CpGs using the IPA software revealed nine canonical pathways with GABA receptor on top of the list and several biofunction categories associated with cellular growth and proliferation that were significantly enriched in CPCs in comparison with CPPs or aCPPs. To identify minimal CPC specific signature, we applied a difference in DNA methylation of 40% and a p-value of 0.001. This increased statistical stringency led to the identification of 59 CpG sites encompassing 33 candidate genes. Of them, 3 genes were validated by pyrosequencing in the initial CPT cohort (n = 34) and a new replication cohort of n = 23 CPT samples. Next, we tested the sensitivity of the CPC specific DNA methylation signature against DNA methylation profiles of several brain tumor datasets extracted from GEO database and found that CPC-DNA methylation signature was highly specific. Our data suggest that dysregulation of epigenetic mechanisms contribute to the molecular events leading to tumor development and progression in CPC and that our DNA methylation based biomarker signature can have prognostic value for this disease and enable better treatment strategies.