Microbial natural products have been a source for new drugs for many decades and are unrivaled in their capacity to generate not only future therapeutic agents, but also providing key agents for agricultural and industrial use. LC-MS/MS based metabolomic tools and technologies have been developed that can rapidly dereplicate nonribosomal peptides and statistically identify related congeners in an automated nontargeted process from complex natural product extracts with nanogram sensitivity. This data-base search approach is designed to handle linear, cyclic and cyclic-branched nonribosomal peptides from proteinogenic and nonproteinogenic amino acids without genomic data or traditional bioactivity directed fractionation. Chemometric work-flows combined with a comprehensive metabolomic guided discovery strategy were used to profile the chemical space of a diverse collection of understudied fungal endophytes from fruiting plants. This approach allowed for the prioritization of unique isolates and for the focused discovery, isolation and characterization of distinct outlier metabolites by LC-SPE, 1D and 2D NMR, HRMS and single crystal X-ray analysis. These metabolomic tools and strategies have led to the discovery and characterization of 35 new and over 40 known natural products, many of which are biologically active. This thesis with enabling metabolomic tools and novel discoveries has demonstrated the utility of these analytical methodologies as an effective strategy for the untargeted discovery of new natural products from bacteria and endophytic fungi. / Thesis / Doctor of Philosophy (PhD)