The quality of fermented soybeans can be determined by diverse metabolites produced by microorganisms. Mass spectrometry-based metabolomic approach was applied to investigate the differences in volatile and non-volatile metabolite profiles of fermented soybeans by different microorganisms [e.g., molds, yeasts, lactic acid bacteria (LAB), and other bacteria]. The partial least squares-discriminant analysis (PLS-DA) for volatile metabolites profiles indicated that the fungi group (mold/yeast) was clearly discriminated from the bacteria group (bacteria/ LAB). The metabolic pathways related to the formation of volatile metabolites also differed according to microorganisms. In particular, the formation of branched-chain aliphatic alcohols and esters increased in the fungi group, while that of volatiles derived from fatty acids was superior in the bacteria group. In addition, we could determine the microorganism-specific metabolites using a correlation network analysis. This study can provide the fundamental knowledge on the metabolic differences according to the type of microorganisms in fermented soybeans.