Nontargeted lipidomics enables both hypothesis-testing studies and discovery-driven projects in biology. However, classic nontargeted lipidomics analyses do not provide accurate concentration levels of the hundreds of annotated lipids, inhibiting comparison and integration of datasets across studies. Improvements in nontargeted lipidomics relies on developing more robust and standardized workflows that can provide reproducible, high-throughput quantification of diverse lipid species. Ultimately, achieving accurate quantification would ensure that results are reproducible and transferable across laboratories, thus enhancing the reliability of lipidomics as a tool for clinical and research applications. Hence, we explore strategies for sample preparation, data curation, and the integration of novel workflows to improve quantitative output in nontargeted analyses. Chapter one evaluates the effectiveness of volumetric microsampling devices for nontargeted lipidomic profiling by liquid chromatography-high resolution tandem mass spectrometry (LC-HRMS/MS). With the promise of accessibility for larger studies, the commercial plasma separation card is thoroughly analyzed by standard method validation protocols and compared dually to dried blood spots and traditional venipuncture plasma. This includes tests of repeatability, recovery, and stability, as well as comparing the overall lipidome coverage offered by each approach. We found that the plasma separation card provides acceptable coverage for its low volume and correlates well to standard plasma responses, ultimately outperforming dried blood spots on nearly all criterion. Despite certain class-dependent limitations, we conclude that plasma separation cards represent the most reliable method for quantification in microsampling applications to lipidomic profiling. Chapter two focuses on the curation of nontargeted lipidomic datasets, specifically in regards to the handling of multiple adduct forms of individual lipid species. Data analysis is traditionally done using the primary or most abundant ion species per analyte. However, certain types of lipids, most notably neutral lipids, will form multiple adduct ions that can vary in relative abundance by up to 70% depending on chemical structure and/or ionization conditions. Here, we systematically evaluate adduct formation trends in diacylglycerols (DAG) across eight different biological matrices. First, the consistency of ratios between four established adduct forms of DAGs were reviewed for inter-sample variability, inter-class variability, and variations between study matrices. Next, several factors were investigated to better understand the variance, including acyl chain length, degree of unsaturation, and signal intensity. Lastly, different combinations of these adduct forms were tried to determine the impact of adduct joining on lipid quantification. Our findings emphasize the need for regular adduct evaluation within data processing methods to properly account for response variations that contribute to quantitative inaccuracies. Chapter three provides a novel approach to absolute quantification within the classical nontargeted lipidomics workflow. Typical single-point calibrations are hindered by mismatched ionization effects between the internal standards and endogenous species, thus leading to quantitative inaccuracies when left uncorrected. Current acquisitions of large-scale data can quantify these effects with minimal intervention to the workflow. To this end, we propose the use of pooled QC dilutions to determine the extent of matrix effects in a given retention time window to better inform the alignment and subsequent matching of internal standards for quantitation. Additionally, we calculate response factors from the intensities of subclass-specific internal standards which can be further extrapolated to correct for large disparities in structure-based ionization efficiencies. This information is then used to inform subclass-specific approaches to internal standard selection for absolute quantification. This method demonstrates greater accuracy when compared to traditional single-point calibrations and can be readily applied to most nontargeted workflows.