Lignocellulosic biohydrogen is a promising renewable energy source that could be a potential alternative to the unsustainable fossil fuel-based energy. Biohydrogen production could be performed by Clostridium thermocellum that is the fastest known cellulose-degrading bacterium. Previous investigations have shown that the co-culture of C. thermocellum JN4 and a non-cellulolytic bacterium Thermoanaerobacterium thermosaccharolyticum GD17 produces more hydrogen than the C. thermocellum JN4 mono-culture, but the mechanism of this improvement is unknown. In this work, we carried out genomic and evolutionary analysis of hydrogenase-coding genes in C. thermocellum and T. thermosaccharolyticum, identifying one Ech-type [NiFe] hydrogenase complex in each species, and, respectively, five and four monomeric or multimeric [FeFe] hydrogenases in the two species. Further transcriptional analysis showed hydrogenase-coding genes in C. thermocellum are regulated by carbon sources, while hydrogenase-coding genes in T. thermosaccharolyticum are not. However, comparison between transcriptional abundance of hydrogenase-coding genes in mono- and co-cultures showed the co-culturing condition leads to transcriptional changes of hydrogenase-coding genes in T. thermosaccharolyticum but not C. thermocellum. Further metabolic analysis showed T. thermosaccharolyticum produces H-2 at a rate 4-12-fold higher than C. thermocellum. These findings lead to the suggestion that the improvement of H-2 production in the co-culture over mono-culture should be attributed to changes in T. thermosaccharolyticum but not C. thermocellum. Further suggestions can be made that C. thermocellum and T. thermosaccharolyticum perform highly specialized tasks in the co-culture, and optimization of the co-culture for more lignocellulosic biohydrogen production should be focused on the improvement of the non-cellulolytic bacterium.