Patients with ulcerative colitis (UC) have a higher risk of developing colorectal cancer (CRC), however, the metabolic shifts during the UC-to-CRC transition remain elusive. In this study, an AOM-DSS-induced three-stage colitis-associated colorectal cancer (CAC) model is constructed and targeted metabolomics analysis and pathway enrichment are performed, uncovering the metabolic changes in this transition. Spatial metabolic trajectories in the "normal-to-normal adjacent tissue (NAT)-to-tumor" transition, and temporal metabolic trajectories in the "colitis-to-dysplasia-to-carcinoma" transition are identified through K-means clustering of 74 spatially and 77 temporally differential metabolites, respectively. The findings reveal two distinct metabolic profile categories during the inflammation-to-cancer progression: those with consistent changes, either increasing (e.g., kynurenic acid, xanthurenic acid) or decreasing (e.g., long-chain fatty acids, LCFAs), and those enriched at specific disease stages (e.g., serotonin). Further analysis of metabolites with consistent temporal trends identifies eicosapentaenoic acid (EPA) as a key metabolite, potentially exerting anti-inflammatory and anti-cancer effects by inhibiting insulin-like growth factor binding protein 5 (IGFBP5). This study reveals novel metabolic mechanisms underlying the transition from UC to CAC and suggests potential targets to delay the progression.