Purpose <p>Ammonia volatilization from nitrogen fertilizers contributes to atmospheric pollution and nitrogen loss in agricultural systems. This study assessed ammonia volatilization and growth patterns of Chinese cabbage using two types of nitrogen fertilizers: uncoated urea (UR) and latex-coated slow-release urea (LCU).</p> Materials and methods <p>Field experiments employing nitrogen rates of 0, 160, and 320&#xa0;kg/ha were conducted to evaluate emissions and crop performance. Ammonia volatilization was monitored using passive air samplers, and machine learning models, including Random Forest (RF) and Gradient Boosting (GBST), were utilized to examine environmental influences.</p> Results and discussion <p>Results indicated that LCU decreased cumulative ammonia volatilization by 12.7% compared to UR, suggesting its efficacy in reducing nitrogen loss and enhancing air quality. Soil temperature and moisture were identified as primary factors influencing ammonia volatilization, facilitating urea hydrolysis and emissions. Machine learning analysis confirmed the significance of these variables, along with fertilizer application rates, in determining ammonia volatilization patterns. Conversely, LCU resulted in lower crop yields than UR, possibly due to its single application method, which may not meet the nutrient requirements of the crop throughout its growth stages. UR yielded higher crop outputs with split applications, demonstrating superior nutrient availability and uptake efficiency but accompanied by greater ammonia emissions.</p> Conclusions <p>This study highlights the capacity of LCU to mitigate ammonia emissions and underscores the necessity to refine its application techniques to enhance crop productivity while minimizing nitrogen losses. These insights advocate for sustainable agricultural practices and refined fertilizer management strategies.</p>