OBJECTIVE: To identify DNA methylation patterns of heavy smokers in oral rinse samples. METHODS: Genome-wide DNA methylation data was imported from Gene Expression Omnibus using the GEOquery package. Two independent sets were analyzed: (a) 71 epigenomes of cancer-free subjects (heavy smokers = 37 vs. non-smokers = 31); for concordance assessment (b) 139 oral-cancer patients' epigenomes (heavy smokers = 92 vs. non-smokers = 47). Differential DNA methylation for CpG positions and at the regional level was determined using Limma and DMRcate Bioconductor packages. The linear model included sex, age, and alcohol consumption. The statistical threshold was set to < 0.05. Functional gene prioritization analysis was performed for gene-targeted analysis. RESULTS: In individuals without cancer and heavy smokers, the gene was found with two CpG positions differentially hypermethylated ( = 0.012 after FDR adjustment), in a region of 48 bp with an absolute methylation difference >10% between groups ( = 1.76 × 10). In the analysis corresponding to oral-cancer patients, we found differentially hypomethylated cancer patients, but also in subjects without oral cancer in the targeted analyses. Remarkably, was found differentially hypermethylated in heavy smokers without a diagnosis of cancer in two consecutive probes cg05575921 ( = 3.13 × 10) and cg10208897 ( = 1.36 × 10). CONCLUSIONS: Differentially methylated ADAMTS2, and genes are biomarker candidates in oral rinse samples.