Objectives <p>This study aimed to investigate the predictive value of contrast-enhanced computed tomography (CECT) characteristics of peritumoral lymph node (LN) combined with composite inflammatory markers in lymph node metastasis (LNM) for patients with T1 colorectal cancer (CRC), and evaluated their diagnostic efficacy.</p> Methods <p>This retrospective study included 212 patients with T1 CRC (non-LNM: <i>n</i> = 185; LNM: <i>n</i> = 27). The CECT characteristics of the peritumoral LN and inflammatory markers were analyzed. Variables with statistical significance were included in penalized logistic regression (the least absolute shrinkage and selection operator, LASSO) for further feature selection and coefficient shrinkage. The diagnostic performance of the prediction models was evaluated using the receiver operating characteristic curve. The net reclassification improvement (NRI) and integrated discrimination improvement (IDI) were calculated to quantify the incremental predictive value between the models, and the decision curve analysis was plotted to evaluate the clinical practicability.</p> Results <p>27/212 (12.74%) patients had LNM. Six CECT characteristics, three inflammatory markers and six clinical pathological parameters differed significantly between the two groups (all <i>p</i> &lt; 0.05). Eight non-zero coefficient variables were retained. The preoperative prediction model based on CECT characteristics and inflammatory markers had the highest AUC (AUC = 0.830), with NRI and IDI both greater than 0, and had the optimal clinical decision-making value.</p> Conclusions <p>Preoperative CECT characteristics of LN and composite inflammatory markers were significantly correlated with the occurrence of LNM in T1 CRC patients. The combined model may be useful for the early identification of LNM.</p>