Abstract The most common complication to affect older adults after surgery is the development of a perioperative neurocognitive disorder. Up to 30% of patients over 60 develop postoperative cognitive dysfunction (POCD), also termed neurocognitive dysfunction (postoperative) (NCD), within 6 weeks after a surgical procedure. POCD/NCD is characterized by impairment of memory, attention, learning, concentration and/or executive function on psychometric testing. Patients with POCD/NCD may experience persistent cognitive dysfunction over 7 years after surgery, greater loss of independence, leaving the labor market, higher health care costs, and increased morbidity and mortality. The etiology and pathogenesis of POCD/NCD are poorly understood. At present, no biomarkers of susceptibility to POCD/NCD are available for use before surgery, or for guiding diagnosis and management of POCD/NCD. Mild cognitive impairment and late onset Alzheimer's disease are risk factors for cognitive decline after surgery suggesting overlapping pathophysiology. Recently, we reported over 450 differentially methylated positions in DNA from blood samples that distinguish persons with Alzheimer's disease from cognitively healthy persons matched for age and sex. Patterns of DNA methylation regulate gene expression by coordinating the influence of environmental factors and genetic coding sequences. Accordingly, in this application we will test whether blood samples acquired before surgery provide DNA methylation biomarkers of Alzheimer's disease that distinguish patients who are at risk for POCD/NCD after surgery from those who are not (Specific Aim 1). As well, we will test blood samples acquired 6 weeks after surgery for biomarkers of Alzheimer's disease that are differentially methylated from baseline levels before surgery in patients with and without POCD/NCD (Specific Aim 2). Our technical innovation is the use of the Illumina Infinium MethylationEPIC BeadChip to classify DNA methylation status at over 850,000 candidate cytosine loci in longitudinal blood samples from patients with and without POCD/NCD. These data will provide new predictors of susceptibility to POCD/NCD for the diagnosis, prognosis, and care of patients with POCD/NCD. In turn, differentially methylated positions at previously unknown loci and pathways will support a more complete understanding of heritable and acquired mechanisms that underlie POCD/NCD with potential to identify novel therapeutic targets.