Endocrine therapy (ET) is one of a number of targeted therapies for estrogen receptor-positive breast cancer (BRCA); however, resistance to ET has become the primary issue affecting treatment outcome. In the present study, a predictive classifier was created using a DNA methylation dataset to identify patients susceptible to endocrine resistance. DNA methylation and RNA sequencing data, and the clinicopathological features of BRCA, were obtained from The Cancer Genome Atlas. Stringent criteria were set to select and classify patients into two groups, namely those resistant to ET (n=11) and sensitive to ET (n=21) groups. Bump hunting analysis revealed that 502 out of 135,418 genomic regions were differentially methylated between these two groups; these regions were differentially methylated regions (DMRs). The majority of the CpG sites contained in the DMRs mapped to the promoter region. Functional enrichment analyses indicated that a total of 562 specific genes encompassing these DMRs were primarily associated with 'biological progress of organ morphogenesis and development' and 'cell-cell adhesion' gene ontologies. Logistic regression and Pearson's correlation analysis were conducted to construct a predictive classifier for distinguishing patients resistant or sensitive to ET. The highest areas under the curve and relatively low Akaike information criterion values were associated with a total of 60 DMRs; a risk score retained from this classifier was revealed to be an unfavorable predictor of survival in two additional independent datasets. Furthermore, the majority of genes (55/63) exhibited a statistically significant association between DNA methylation and mRNA expression (P<0.05). The association between the mRNA expression of a number of genes (namely calcium release activated channel regulator 2A, Schlafen family member 12, chromosome 3 open reading frame 18, zinc finger protein 880, dual oxidase 1, major histocompatibility complex, class II, DP β1, C-terminal binding protein 1, ALG13 UDP-N-acetylglucosaminyltransferase subunit and RAS protein activator like 2) and the prognosis of patients with estrogen receptor-positive BRCA and ET resistance was determined using Kaplan-Meier Plotter. In summary, the predictive classifier proposed in the present study may aid the identification of patients sensitive or resistant to ET, and numerous genes maybe potential therapeutic targets to delay the development of resistance to ET.