Site-specific calibration of laser induced breakdown spectroscopy (LIBS) analysis can be achieved by presenting known geological sample data to update algorithms, therefore improving analysis of unknown samples post-calibration. The goal of this project is to optimise drill core analysis by LIBS for samples from Kittilä gold mine. In total 227 samples have been produced and analysed with a range of analytical techniques to build a comprehensive, site-specific reference library. The data from these analyses has been input into the reference library and utilised by algorithms for spectral peak identification during LIBS analysis of bulk drill core material. The semi-quantitative and quantitative data of 40 Kittilä samples has been used as a comparison to ICP-AES data used by the mine. The Kittilä powders were used to produce 39 pellets, each representative of a section of drill core of which the ICP-AES and TXRF data was known. The pellets were scanned by LIBS and the data then compared to the known data. LIBS showed strong correlations (>0.75) with ICP-AES data for As, Cr, Cu, Fe, Mn and S, poor correlations were seen for Pb and Co, however these can be explained by issues related to ICP-AES. LIBS pellet data proved strongly competitive as 5/9 elements displayed stronger correlations between LIBS and ICP-AES than TXRF. Subsequently, the successful scanning of drill core boxes directly has proven that calibration can increase the speed with which the bulk material can be effectively analysed.