The analysis of spectroscopic data to solve chemical structures requires practical skills and drills. In this context, we have developed ULg Spectra, a computer-based tool designed to improve the ability of learners to perform complex reasoning. The identification of organic chemical compounds involves gathering and interpreting complementary information from mass, infrared, Raman, and nuclear magnetic resonance spectra. Here, special attention is paid to one-dimensional H-1 and C-13 NMR spectra and to two-dimensional NMR spectra because these techniques particularly require extensive interactive data manipulation. ULg Spectra offers tutorial-drill materials, including spectra that are "authentic" in the sense that they contain solvent and impurity traces rather than being "idealized" spectra often found in textbook examples. A public version is accessible online free of charge. The exam results for two groups of students, one having used ULg Spectra for extra home-based training and the other not, were compared. Statistical data demonstrated higher performance for the "trained" students compared with the control group.