Primary Age Related Tauopathy (PART) and Alzheimer’s disease (AD) are characterized by the accumulation of intracellular phosphorylated tau (p-tau) aggregates, also known as neurofibrillary tangles (NFTs), in temporo-limbic brain regions. In PART, NFTs remain confined to the hippocampus whereas in AD they spread to neocortical regions. Furthermore, individuals with PART do not develop the extracellular ß-amyloid neuritic plaques characteristic of AD and generally show less neurodegeneration and cognitive decline. Thus, PART presents a unique opportunity to study mechanisms of p-tau accumulation independent of ß-amyloid co-pathology, and may represent a model of AD resistance. Previous studies have identified genetic modulators of PART, but there remains a gap in our knowledge of the additional mechanisms predictive of PART vs. AD phenotypes during aging. DNA methylation (DNAm) changes with aging, lifestyle, and environmental exposures and has been used to develop biomarkers for biological aging and mortality risk. To test the hypothesis that DNAm drives variation in p-tau pathology in PART and may regulate the transition to AD, we conducted an epigenome wide association study in a neuropathologically confirmed cohort of individuals with PART (n=260). 13 CpG sites associated with hippocampal p-tau, and integrative transcriptomic analysis identified genes involved in synaptic transmission processes. We leveraged a second cohort (n=707) spanning the full spectrum of PART-AD neuropathology and developed DNAm-based machine learning models to predict AD status and regional p-tau burden in hippocampus and frontal cortex. Models predictive of frontal cortical p-tau associated with neuroinflammation, while the PART-AD classifier tracked metabolic processes and stratified a subset of indeterminate cases into distinct groups with significantly different levels of cognitive impairment. Finally, to address the need for more scalable epigenetic technologies for future EWAS studies in PART and AD, we simultaneously developed a novel microarray for high-throughput population-scale screening of DNA methylation. The designed array successfully deconvolved cell types from diverse human tissues and identified age-related alterations in DNA methylation and its oxidized derivative, 5-hydroxymethylation. Collectively, this work identified distinct epigenetic signatures associated with tau pathology and PART vs. AD phenotypes, while producing a novel technology for future epigenetic investigations into brain health and disease.