Alzheimer's disease (AD) is the most common neurodegenerative disease causing dementia in more than 5.4 million individuals in the United States. AD is a complex disease with a large genetic factor. Genetic linkage studies identified three causal genes for the Mendelian form of AD, and genome-wide association studies (GWAS) have discovered more than 25 genetic loci associated with AD risk. However, it is unclear how these loci influence AD, or if they are associated with other AD phenotypes such as age at onset (AAO) or disease progression. Loci identified thus far explain less than half of the estimated genetic influence on risk, suggesting there are many more to be identified. Genetic studies of endophenotypes, biomarkers that are genetically correlated with disease and part of the causal pathway, have successfully identified novel genetic loci associated with AD risk and other phenotypes. We hypothesized that genetic studies using potential and well-validated endophenotypes may reveal novel associations with disease and provide important information about biology underlying the genetic architecture shared between the endophenotype and disease, not only with risk but also disease modifiers such as AAO or rate of cognitive decline. We also hypothesized that we could identify novel endophenotypes by unbiased genetic studies of a variety of biologically relevant protein levels.To identify plasma proteins that may be biomarkers for complex traits, including AD, we performed single-variant analyses on 818 individuals with levels of 146 plasma proteins measured using a multiplex immunoassay. This multiplex assay included a wide variety of proteins such as markers of various diseases and biological mechanisms. Although we did not identify novel AD biomarkers, we did identify promising novel biomarkers for multiple sclerosis and stroke, among others. We also observed evidence of complex genetic regulation of several proteins including potential pleiotropy and loci with multiple variants independently influencing protein levels.Several putative CSF biomarkers for AD have emerged the past few years, including angiotensin-converting enzyme (ACE), clusterin (CLU), and chitinase-3-like protein 1 (also known as YKL-40, encoded by CHI3L1). We wanted to determine whether these potential biomarkers could also be AD endophenotypes; therefore, we analyzed these proteins in addition to known endophenotypes for AD: CSF amyloid-beta (A[beta]42), phosphorylated tau (ptau181), tau, and apolipoprotein E (ApoE). ACE, ApoE, and YKL-40 levels were associated with the respective protein encoding genetic loci (ACE, APOE, and CHI3L1 respectively). The top loci for ACE and ApoE were also associated with AD risk and AAO, suggesting these may be informative for pathogenesis influencing disease onset. We did not find evidence that YKL-40 is an AD endophenotype, but CSF levels of YKL-40 positively correlated with both tau and ptau181 and the strength of the correlation significantly increased after including the genetic information from the associated locus. This suggests that biomarker studies may benefit by including genetic information. Our genetic analyses of CSF tau, ptau181, and A[beta]42 replicated results published previously from a smaller cohort and revealed two novel loci associated with ptau181. For the first time, we also identified two loci outside the APOE region that were associated with A[beta]42 which were also associated with AD risk (GLIS1 on 1p32.3: [beta] = 0.100, P = 3.43x10−2), disease progression (GLIS1: [beta] = 0.277, P = 1.92x10−2), and AAO (SERPINB1 on 6p25: [beta] = 0.043, P = 4.62x10−3).Most of the loci identified were non-coding; therefore, requiring additional analyses to determine the functional genetic traits affecting protein levels and potentially influencing AD phenotypes. We analyzed CSF levels of ACE and ApoE with available exome-chip and whole-genome data to identify coding variants. ApoE levels appeared to be driven by APOE genotype, there were no additional variants associated with ApoE levels. Several variants for ptau181 and A[beta]42 were predicted to alter regulatory motifs and affect protein binding. In order to ascertain the functional genes potentially driving the genetic association, we searched for SNPs with expression quantitative trait locus effects in human tissues. Analyses indicated that the same variant on 6p25 influenced A[beta]42 levels and expression of SERPINB1, which is a key regulator for neutrophil activity. Recently A[beta]42 was reported necessary to trigger infiltration of neutrophils into the brain, and with our findings, adds to the growing evidence that immune response plays a key role in AD pathogenicity. By utilizing quantitative traits such as endophenotypes in genetic studies we obtained results that will inform future studies.