Proton (1H) NMR spectroscopy presents a powerful tool for biomass mixture studies by revealing the involved chemical compounds with identified ingredients and molecular structures. However, conventional 1H NMR generally suffers from spectral congestion when measuring biomass mixtures, particularly biomass carbohydrate samples, that contain various physically and chemically similar compounds. In this study, a targeted detection NMR approach, DREAMTIME, is exploited for studying biomass carbohydrate mixtures by spectroscopically targeting the desired compounds in separate 1D NMR spectra. From three model mixtures, namely, the C6 sugar isomerization mixture, C5 sugar catalytic hydrogenation mixture, and d-glucose and d-xylose fermentation mixture, to a real reaction mixture of sucrose hydrolysis, DREAMTIME achieves satisfactory performance in compound identification and mixture analysis even when mixture signals are crowded and overlapped in conventional 1H NMR. Additionally, DREAMTIME is performed in a rapid 1D acquisition manner, making it available for highly efficient analysis on various biomass reactions. Our results suggest that DREAMTIME provides an effective method for wide applications to complex biomass mixture analysis with explicit compound identification, targeted component monitoring, and conversion reaction detection.