This study aimed to apply high-resolution metabolomics to detect compounds that may contribute significantly to prostate cancer (PCa) development. The test population's sera for evaluating the metabolic differences consisted of healthy control (n = 96) and PCa (n = 50) groups. PCa patients were further divided into two groups based on whether their PSA level was >4 (n = 25) or <4 (n = 25). Univariate analysis was performed with the false discovery rate (FDR) at q = 0.05 to determine significantly different metabolites. Principal component analysis (PCA) and hierarchical clustering analysis (HCA) clearly distinguished healthy subjects from PCa groups, while no significant difference was observed in PCa patients with PSA level < 4 or > 4. Mummichog, in combination with the KEGG and MetaboAnalyst, showed that tryptophan metabolism along the kynurenine pathway was most significantly enriched, with −log (p) < 0.05. L-Tryptophan, kynurenine, anthranilate, isophenoxazine, glutaryl-CoA, (S)-3-hydroxybutanoyl-CoA, acetoacetyl-CoA, and acetyl-CoA were upregulated in correlation with the PSA level of PCa patients; in contrast, indoxyl, indolelactate, and indole-3-ethanol, involved in the alternative pathway, were downregulated in the PCa patients. Validation and quantification of potential metabolites by MS/MS further confirmed the disruption of tryptophan, kynurenine, and anthranilate, suggesting that the metabolites of this pathway are potential biomarkers in patients with PCa.