Urban rivers dynamically interfered by anthropogenic activities are considered as a vital reservoir of antibiotic resistance genes (ARGs). Here, a total of 198 ARGs and 12 mobile genetic elements (MGEs) were profiled in water and sediment from the Chaobai river, Beijing. The total abundances of ARGs (1.01 × 10-4.58 × 10 copies/L in water and 2.92 × 10-3.34 × 10 copies/g in sediment), which were dominated by beta-lactamase genes, exhibited significant seasonal variations (p < 0.05). Significant linear correlations between the total abundances of ARGs and MGEs were observed in both water and sediment (p < 0.01). Variance partitioning analysis disclosed that environmental variables (i.e., water temperature (WT), dissolved oxygen (DO), nutrients, metals, etc.) and antibiotics were the main contributors to the variations of ARGs and MGEs, and explained 55-80 % and 27-67 % of the total variations in ARGs and MGEs, respectively. The partial least-squares path model revealed the ARG abundances in water and sediment were affected by environmental variables and antibiotics both directly and indirectly but by MGEs directly. Moreover, random forest algorithm explored that WT, Ni, DO, Co, and polyether and macrolide antibiotics were the main drivers (>10 %) of ARGs dissemination in water, whereas the transposase genes of Tp614, tnpA, and IS613 were the main drivers of ARGs dissemination in both water and sediment. This study provides a comprehensive understanding of the driving factors for the ARGs dissemination in an urban river, which is of great significance for risk management of antibiotic resistome.