Colitis-associated colorectal cancer (CAC) is fatal and can develop spontaneously or as a complication of inflammatory bowel diseases. Although co-administration of azoxymethane/dextran sulfate sodium (AOM/DSS) is a classic method for CAC modeling, its limitations need to be addressed. Accordingly, we aimed to optimize the AOM/DSS model to study CAC extensively and further investigate its pathogenic mechanisms relative to microbiota and metabolism. We optimized the CAC model via a single or enhanced injection of AOM combined with different administration modes and varying DSS concentrations. Subsequently, the fecal-microbiota composition was examined using 16S RNA sequencing, and fecal-colon-metabolome profiles were evaluated via ultra-high performance liquid chromatography-mass spectrometry. Two interval injections of AOM combined with 1.5 % DSS-free drinking resulted in a high tumor formation rate, uniform tumor formation, and low mortality. Based on this model, we innovatively divided the pathogenesis of CAC into three stages, namely inflammation induction, proliferation initiation, and tumorigenesis, and examined the pathological characteristics in each stage. Gut microbial dysbiosis and metabolic alteration drove colorectal tumorigenesis by aggravating inflammation while promoting cell proliferation and carcinogenesis in mice. For the first time, we dynamically demonstrated the process of colon "inflammation to cancer" transformation and provided novel insights to clarify the role of amino acid metabolism in the formation of CAC.