Human tumors are comprised of heterogeneous cell populations that display diverse molecular and phenotypic features. To examine the extent to which epigenetic differences contribute to intratumoral cellular heterogeneity, we have developed a high-throughput method, termed MAPit-patch. The method uses multiplexed amplification of targeted sequences from submicrogram quantities of genomic DNA followed by next generation bisulfite sequencing. This provides highly scalable and simultaneous mapping of chromatin accessibility and DNA methylation on single molecules at high resolution. Long sequencing reads from targeted regions maintain the structural integrity of epigenetic information and provide substantial depth of coverage, detecting for the first time minority subpopulations of epigenetic configurations formerly obscured by existing genome-wide and population-ensemble methodologies. Analyzing a cohort of 71 promoters of genes with exons commonly mutated in cancer, MAPit-patch uncovered several differentially accessible and methylated promoters that are associated with altered gene expression between neural stem cell (NSC) and glioblastoma (GBM) cell populations. In addition, considering each promoter individually, substantial epigenetic heterogeneity was observed across the sequenced molecules, indicating the presence of epigenetically distinct cellular subpopulations. At the divergent MLH1/EPM2AIP1 promoter, a locus with three well-defined, nucleosome-depleted regions (NDRs), a fraction of promoter copies with inaccessible chromatin was detected and enriched upon selection of temozolomide-tolerant GBM cells. These results illustrate the biological relevance of epigenetically distinct subpopulations that in part underlie the phenotypic heterogeneity of tumor cell populations. Furthermore, these findings show that alterations in chromatin accessibility without accompanying changes in DNA methylation may constitute a novel class of epigenetic biomarker.