Newcastle disease virus (NDV) is a significant poultry pathogen, causing ongoing economic losses in China’s poultry industry. To understand its circulation and risk factors associated with farm infection, we carried out an epidemiological study on 1,065 farms across 18 provinces from 2019 and 2022. We collected a total of 43,696 swabs and tested them for NDV using an RT-PCR that targets the partial F gene. The overall bird-level NDV prevalence in the 4 years (2019, 2020, 2021, and 2022) were 1.49% (95%CI: 1.27, 1.71%), 1.24% (95%CI: 1.04, 1.44%), 0.59% (95%CI: 0.45, 0.74%), and 0.44% (95%CI: 0.31, 0.58%) respectively, while the farm-level prevalence of the 4 years were 11.27% (95%CI: 7.93, 14.60%), 13.85% (95%CI: 10.10, 17.61%), 12.63% (95%CI: 8.00, 17.26%), and 10.71% (95%CI: 6.38, 15.04%), respectively. The result also showed a high degree of variation in farm-level prevalence (ranging from 0 to 90%) across the provinces. A multiple logistic regression analysis revealed that four factors, namely layer farm (OR = 2.282, 95% CI: 1.211, 4.298), broiler farm (OR = 2.339, 95% CI: 1.206, 4.534), using mixed feed (OR = 2.973, 95% CI: 1.962, 4.505), and indoor housing with some free-range access (OR = 1.788, 95%CI: 1.096, 2.917), increased the risk of NDV infection on farms. We isolated 104 NDVs, which were all classified into Class II by the phylogenetic analysis, but from two genotypes: four belonged to genotype I, while the remaining 100 mainly belonged to genotype II, indicating that the circulating NDVs were primarily LaSota-like low pathogenic viruses. We used random forest algorithm to established an occurrence risk model, The accuracy of the model was 90.81%. This large-scale investigation revealed NDV prevalence at bird, farm, and province levels over the 4 years. It also identified risk factors associated with farm NDV infections. The findings provide new insights into the epidemiology of NDV in China and offer references for global NDV control.