A process was developed to ascertain the bioactive components of black raspberry (Rubus occidentalis L.) fruit extracts by relating chemical constituents determined by high-field nuclear magnetic resonance (NMR) spectroscopy to biological responses using partial least-squares regression analysis. To validate our approach, we outlined relationships between phenolic signals in NMR spectra and chemical data for total monomeric anthocyanin (TMA) content and antioxidant capacity by the ferric-reducing antioxidant power (FRAP) and 2,2-diphenyl-1-picrylhydrazyl (DPPH) assays. Anthocyanins, cyanidin 3-O-rutinoside (Cy 3-rut), cyanidin 3-O-(2(G))-xylosylrutinoside (Cy 3-xylrut), and cyanidin 3-O-glucoside (Cy 3-glc), were significant contributors to the variability in assay results, with the two most important NMR bins corresponding to the methyl peaks in Cy 3-rut (6''') and/or Cy 3-xylrut (6(IV)). Many statistically important bins were common among assay models, but differences in structure-activity relationships resulted in changes in bin ranking. The specificity of these results supported the application of the process to investigate relationships among health-beneficial natural products and potential biological activity.