High levels of nitrogen originating from blasting operations, for example at mining sites or quarries, risk contaminating water bodies through leaching from waste rock dumps. Woodchip bioreactors can be a simple and cost-effective way of reducing nitrate concentrations in the leachate. In this study we investigated how bottle sedge, barley straw, and pine woodchips used as electron donors for denitrification influenced microbial community composition and nitrate removal in lab-scale bioreactors during 270 days. The reactors were operated to ensure that nitrate was never limiting and to achieve similar nitrate removal (%). Distinct bacterial communities developed due to the different substrates, as determined by sequencing of the 16S rRNA gene. Sedge and straw reactors shared more taxa with each other than with woodchips and throughout the experimental period, sedge and straw were more diverse than woodchips. Cellulose degrading bacteria like Fibrobacteres and Verrucomicrobia were detected in the substrates after 100-150 days of operation. Nitrate removal rates were highest in the sedge and straw reactors. After initial fluctuations, these reactors removed 5.1 – 6.3 g N m^-3 water day^-1, which was 3.3 – 4.4 times more than in the woodchip reactors. This corresponded to 48 %, 42 %, and 44 % nitrate removal for the sedge, straw, and woodchip reactors respectively. The functional communities were characterized by quantitative PCR and denitrification was the major nitrate removing process based on genetic potential and water chemistry, although sedge and straw developed a capacity for ammonification. Gene ratios suggested that denitrification was initially incomplete and terminating with nitrous oxide. An increase in abundances of nitrous oxide reducing capacity in all substrate types towards the end increased the potential for less emissions of the greenhouse gas nitrous oxide.