Osteomyelitis, which often arises from a surgical-site infection, is a serious problem in orthopaedic surgery. However, there are no specific biomarkers for osteomyelitis. Here, to identify specific plasma biomarkers for osteomyelitis, we conducted metabolome analyses using a mouse osteomyelitis model and bioluminescence imaging. We divided adult male pathogen-free BALB/C mice into control, sham-control, and infected groups. In the infected group, a bioluminescent Staphylococcus aureus strain was inoculated into the femur, and osteomyelitis was detected by bioluminescence imaging. We next analysed the metabolome, by comprehensively measuring all of the small molecules. This analysis identified 279 metabolites, 12 of which were significantly higher and 45 were significantly lower in the infected group than in the sham-control and control groups. Principal component analysis identified sphingosine as the highest loading factor. Several acyl carnitines and fatty acids, particularly ω-3 and ω-6 polyunsaturated fatty acids, were significantly lower in the infected group. Several metabolites in the tricarboxylic acid cycle were lower in the infected group than in the other groups. Thus, we identified two sphingolipids, sphinganine and sphingosine, as positive biomarkers for mouse osteomyelitis, and two components in the tricarboxylic acid cycle, two-oxoglutarate and succinic acid, as negative biomarkers.