Alzheimer's disease (AD) is a neurodegenerative disease, the most common cause of dementia in elderly persons. Accumulation of amyloid plaques in the brain is a characteristic of AD. The requirements for a biomarker include the ability to measure a pathologic process, predict outcome, distinguish disease or measure a pharmacological response to a drug treatment or therapeutic intervention. However, there are no reliable blood based biomarkers for AD. Serum-based protein measurement is a routine practice for biomarkers in human disease, but comprehensive profiling of serum proteome is often masked by the twenty most abundant proteins and impacted by large dynamic range. The commonly used depletion method can alleviate the challenge but introduce additional experimental variation. Here we present a deep analysis of un-depleted human serum proteome by combining 11-plex TMT labeling, exhaustive 2D liquid chromatography fractionation, and high resolution tandem mass spectrometry. This platform is capable of identifying 4,500 protein components, covering 6 orders of dynamic range, representing one of the deepest serum proteome datasets. Finally, a subset of proteins, show statistically significant difference between AD and control samples, which may serve as biomarker candidates.