A biological denitrifying process was employed for the treatment of nanofiltration (NF) concentrate with high conductivity, which was generated from coking wastewater in a sequencing batch reactor (SBR). The results showed that the average removal efficiencies of chemical oxygen demand (COD), total nitrogen (TN) and nitrate were 47.6%, 61.1% and 94.6%, respectively. Different microbial communities were identified by sequencing the V1-V3 region of the 16S rRNA gene using the MiSeq platform, showing that the most abundant bacterial phylum in the SBR system was Proteobacteria, with the subclasses β-Proteobacteria and α-Proteobacteria being dominant. The key microorganisms responsible for denitrification belonged to the genera Thauera, Hyphomicrobium, Methyloversatilis, Hydrogenophaga, Ignavibacterium, Rubrivivax and Parvibaculum. Quantitative real-time polymerase chain reaction was used to assess the absolute abundance of microbial genera, using 16S rRNAs and denitrifying genes such as narG, nirS, nirK, nosZ, in both SBR start-up and stable operation. The abundances of narG, nirK and nosZ were lower during stable operation than those during the start-up period. The abundance of nirS at a level of 104–105copies/ng in DNA was much higher than that of nirK, thus being the dominant functional gene in nitrite reduction.