This study aimed to develop a predictive nomogram model and a risk classification system to predict the likelihood of lymph node metastases for non-metastatic muscle-invasive bladder cancer (MIBC) patients using a large population-based cancer database. According to our nomogram, larger tumor size, overlapping lesions, young age, female, poorly differentiated histological grade, and advanced T stage, are independent risk factors for pN+. A precise nomogram model predicting pN+ probability for MIBC patients can support patient risk stratification and outcome estimation, and eventually guide clinical decision-making.