Abstract Woodchip bioreactors are increasingly used as tools to mitigate nitrogen (N) pollution from agricultural drainage water. They consist of a basin filled with woodchip material through which N contaminated drainage water can flow. During the water transport through the filter matrix, oxygen is rapidly depleted and denitrification removes a fraction of the nitrate N present in the water. However, the N removal efficiency of the bioreactors varies significantly both across systems and seasonally. Furthermore, denitrification can also produce nitrous oxide, which is a potent greenhouse gas. Here, we investigated how variation in hydraulic residence time influenced N removal efficiency and nitrous oxide emissions at eight woodchip bioreactors of different flow designs, monitored for 2–4 years. We also characterised the relative abundance of genes involved in the N cycle at three of the bioreactors using metagenomics. Our results showed that total N removal was 17–73% of the yearly incoming N and that it was influenced by hydraulic residence time and water temperature. Nitrous oxide emissions were variable among the different bioreactors and were higher when the hydraulic residence time was less than 60 h. However, the yearly nitrous oxide release did not exceed 2.4% of the nitrate removal (on N atom basis) and the mean among the bioreactors was 0.6%. Although there were marked differences in nitrate removal and nitrous oxide emissions, there were no clear differences in the relative abundance of N-cycling genes among and within three tested bioreactors. Yet, denitrification genes greatly outnumbered genes related to dissimilatory nitrate reduction to ammonium. Overall, our study showed that all eight bioreactors were effective in removing N from agricultural drainage water and that nitrous oxide emissions were low, especially at hydraulic residence times of 60 h or more.