Abstract Cervical lymph node metastasis of esophageal squamous cell carcinoma (ESCC) is an important basis for lymph node dissection of ESCC. Now, there are still disputes about whether to perform two field lymph node dissection or three field lymph node dissection. Establishing prediction model by analysis big data could promote the accuracy of prediction of cervical lymph node metastasis in patients with ESCC which can help clinicians to select individualized surgical scheme for patients. The clinical data of 102 consecutive patients with ESCC who underwent radical esophagectomy and three-field lymph node dissection between December 2015 to December 2021 were retrospectively analyzed. They were divided into positive group and negative group according to cervical lymph node metastasis. The differences between the two groups were compared, and after the univariate analysis those who were statistically significant were included in multivariable logistic regression analysis to establish the risk prediction model and plot a nomogram by software R. Among the 102 patients, univariate analysis(P<0.2)indicated that cervical lymph node metastasis of ESCC was related to the distance from upper edge of the tumor to the incisors (P = 0.101), vascular invasion (P = 0.144) and recurrent laryngeal nerve lymph nodes metastasis (P = 0.159). Multivariate logistic regression analysis showed that the distance from upper edge of the tumor to the incisors (P = 0.032) and the vascular invasion (P = 0.147) were independent risk factors. The AUC of ROC curve is 0.6817 (95% confidence interval: 0.5775–0.7859) and the Youden index is 0.294, 95% CI (0.620, 0.742). The prediction model includes two variables: the distance from upper edge of the tumor to the incisors and the vascular invasion. It plays a certain role in predicting the risk of cervical lymph node metastasis of ESCC, and can assist doctors in individualized treatment for patients with ESCC.