Over the past two decades, metabolomics as a technique has moved from the primary domain of analytical chemists to more widespread acceptance by biologists, clinicians and bioinformaticians alike. Metabolomics offers systems-level insights into the critical roles small molecules play in routine cellular processes and myriad disease states. However, certain unique analytical challenges remain prominent in metabolomics as compared to the other ‘omics sciences. These include the difficulty of identifying unknown features in untargeted metabolomics data, and challenges maintaining reliable quantitation within lengthy studies that may span multiple laboratories. Unlike genomics and transcriptomics data in which nearly every quantifiable feature is confidently identified as a matter of course, in typical untargeted metabolomics studies over 80% of features are frequently not mapped to a specific chemical compound. Further, although many metabolomics studies have begun to stretch over a timeframe of years, data quantitation and normalization strategies have not always kept up with the requirements for such large studies. Fortunately, both experimental and computational strategies are emerging to tackle these long-standing challenges. We will report on several techniques in development in our laboratory, ranging from chromatographic fractionation and high-sensitivity data acquisition, to computational strategies to aid in tandem mass spectrometric spectral interpretation. These developments serve to facilitate analysis for both experts and novice users, which should ultimately help improve the biological insight and impact gained from metabolomics data.