Metabolomics has become an increasingly popular “omics” approach to assist with the robust phenotypic characterization of humans, plants and model organisms. However, the application of metabolomics to livestock phenotyping still lags far behind other applications. To address this shortfall, we have undertaken two major tasks; 1) detailed experimental characterization of the sheep metabolome, and 2) compiling metabolomic data for 5 of the most common livestock species: cattle, sheep, goats, horses and pigs to create the Livestock Metabolome Database [LMDB] (available at www.lmdb.ca). The LMDB is expected to serve as a general hub to support metabolomic studies in livestock. With regard to our work on the sheep metabolome, three experimental projects are underway to look at how metabolomics can be used to assess feed efficiency and carcass merit, pregnancy and litter size, and parasitic infection. The main objectives are to: 1) profile (identify and quantify) the blood metabolome of healthy sheep at different production levels, and 2) identify biomarkers that are predictive for productivity, pregnancy and/or parasite infection status. Our experimental results will be used to expand the LMDB (which already has 1070 metabolites from many other species) and to help develop a “pen-side” metabolomic device to improve sheep selection or management. Preliminary laboratory analyses using nuclear magnetic resonance (NMR) spectroscopy to evaluate serum samples (n=35) collected from lambs and ewes has been promising. We have identified candidate biomarkers that distinguish between lambs with high and low feed efficiency. Metabolite signatures have also been identified that distinguish open and pregnant ewes with a single fetus or multiple fetuses, and parasite-infected sheep. We are currently investigating a larger cohort of samples (n=1720) using 3 different analytical platforms (NMR, ICP-MS, DFI-MS) to validate our initial results.