Extremely metal-poor stars (EMPs, [Fe/H] < -3) are rare, and have shown a variety of interesting element ratios in previous studies, interpreted as due to variations in their supernovae progenitors. In this Gemini LLP, we request 150 hours over five semesters for GRACES high-resolution spectroscopic observations of ~80 EMP stars found in the Pristine survey (Starkenburg et al. 2017). High-resolution (R=40,000), high signal-to-noise (SNR>100) spectroscopy is now necessary to determine accurate chemical abundances for important elements that explore the variety of nucleosynthetic pathways in metal-poor stars, core-collapse supernovae, and the first stars. Sample elements include the alpha (Mg, Si, Ca, Ti), odd-Z (Na, K, Mn), and especially the neutron capture (Sr, Ba) elements, which are sensitive to progenitor mass, metallicity, and rotation rates, supernovae explosion energies, mixing, and fallback, and explosion symmetries or progenitor binarity. Element abundances are also necessary to identify rare classes of chemically peculiar stars, such as CEMP-no, alpha-challenged, and r-process rich stars, which are currently linked to specific epochs or events in galactic chemical evolution scenarios. This program will target EMP stars with 15 < V < 16.5, which is the magnitude range of our on-going follow- up medium-resolution spectral survey at the 2.5-meter Isaac Newton Telescope, and used to confirm their extremely low metallicities and provide carbon abundances. As an LLP, we will increase the number of EMP stars with detailed abundance analyses by ~30%, which is a pivotal leap in sample size for these important observational constraints, and with the potential to explore any previous observational bias due to the more narrow CaHK filter being used by the Pristine survey. It is exciting to realize that any one of these new targets could be the next most metal-poor star (in total metals), providing critical and new constraints on the characterization of the first stars and first supernovae.