Although seasonality is a key driver of environmental fluctuation in aquatic ecosystems, there exists limited knowledge on the factors controlling the distribution of antibiotic resistance genes (ARGs) across seasons at a watershed scale. Here we used high-throughput quantitative PCR to quantify 285 ARGs conferring resistance to most major classes of antibiotics, reveal their spatial and seasonal distribution patterns, and depict the underlying mechanisms in a subtropical riverine ecosystem under low and high human pressures, in Xiamen city, southeast China. Our results showed that spatial differences in ARG richness and abundance overwhelmed their seasonal variations, with only ARGs that confer resistance to sulfonamide and vancomycin being significantly different across seasons. Only a few abundant ARGs (19 ARGs) could contribute to >70% of the total ARGs abundance and were found in all seasons. The significantly higher number of ARGs in the summer rainy period than other seasons coincided with high number of significant edges in ARC co-occurrence networks. Summer rainfall had strong dilution effect on ARCS in upstream waters and enrichment effect in downstream waters. The variance partitioning analysis indicated that the environment explained larger variance of ARG profiles than mobile genetic elements (MGEs), spatial predictors and the rainfall. Nevertheless, strong and significant correlations between transposase gene absolute abundance and aminoglycoside, chloramphenicol, MIS, multi drug and tetracycline classes of resistance genes inferred the role of MGEs on ARG distribution. Overall, our results imply that the modelling and management of ARGs in highly dynamic ecosystems could be better implemented by considering priority genes that dominate at spatial and seasonal gradients.