Prostate cancer (PCa) is a hormone-dependent tumor characterized by an extremely heterogeneous prognosis. Despite recent advances in partially uncovering some of the biological processes involved in its progression, there is still an urgent need for identifying more accurate and specific prognostic procedures to differentiate between disease stages. In this context, targeted approaches, focused on mapping dysregulated metabolic pathways, could play a critical role in identifying the mechanisms driving tumorigenesis and metastasis. In this study, a targeted analysis of the nuclear magnetic resonance-based metabolomic profile of PCa patients with different tumor grades, guided by transcriptomics profiles associated with their stages, was performed. Serum and urine samples were collected from 73 PCa patients. Samples were classified according to their Gleason score (GS) into low-GS (GS < 7) and high-GS PCa (GS >= 7) groups. A total of 36 metabolic pathways were found to be dysregulated in the comparison between different PCa grades. Particularly, the levels of glucose, glycine and 1-methlynicotinamide, metabolites involved in energy metabolism and nucleotide synthesis were significantly altered between both groups of patients. These results underscore the potential of targeted metabolomic profiling to characterize relevant metabolic changes involved in the progression of this neoplastic process.