At present, the clinical utility of metabolomic profiles of human prostate tissue relies on the establishment of correlations between metabolite data and clinical measurements, particularly pathological findings. Because metabolomics is a quantitative study, its clinical value can be rigorously investigated by determining its association with other quantitative measures. The human visual assessment of prostate tissue, however, introduces both inter- and intra-observer biases that may limit the reliability of its quantitations, and therefore, the strength of its correlations with metabolomic profiles. The aim of this study was to develop a simple, feasible protocol for the computer-aided image analysis (CAIA) of prostate pathology slides in order to achieve quantitative pathology from tissue samples, following metabolomic measurement with high-resolution magic angle spinning (HRMAS) magnetic resonance spectroscopy (MRS). Thirty-eight samples from 29 prostatectomy cases were studied with HRMAS MRS. After spectroscopy analysis, samples were serial-sectioned, stained and visually assessed by pathologists. Cross-sections from these samples were then measured with the CAIA protocol. Results showed a two-fold difference between human visual assessments of the area percentages of tissue pathologies and CAIA area percentages obtained for the same features. Linear correlations were found between both metabolites indicative of normal epithelium and those indicative of prostate cancer, and the CAIA quantitative results. CAIA based quantitative pathology is more reliable than human visual assessment in establishing correlations useful for disease diagnosis between prostate pathology and metabolite concentrations.