The methanogenic activity is an important indicator to assess the efficiency of high-solid anaerobic digestion. However, it is not yet elucidated clearly how to detect the parameter rapidly and reliably in the rice straw feeding reactor. Co-inoculated with ruminal digesta and anaerobic sludge, the digestion performance was studied at three different organic loading rates (OLRs). The excitation emission matrix–parallel factor analysis (EEM–PARAFAC) was used to detect dynamic changes in the characteristic of fluorescence components. Our results revealed that CH4 productivity reached 280.90 mL/g volatile solid (VS) with a 54.39% CH4 content under the OLR of 2.26 g/(L⋅d), which amount to 80.29% of its theoretical value. At the OLR of 2.47 g/(L⋅d), the average accumulated NH4+ concentration was 1082.63 mg/L, which resulted in the hydrogenotrophic Methanobacteriales decreasing from 1.70 × 109 to 1.04 × 106 copies/g in the solid residues, whereas the acetotrophic Methanosarcinales increased from 7.89 × 106 to 9.44 × 106 copies/g. The dynamics of the methanogenic community consequently influenced the bioconversion efficiency of rice straw, and CH4 productivity was reduced to 256.54 mL/g VS. The three fluorescent components, at the excitation/emission wavelength of 420 nm/470 nm, 340 nm/430 nm, and 280 nm/340 nm, were decomposed by PARAFAC model in the digestate. Fluorescence intensities of coenzyme F420 and NADH reflected the dynamic changes of CH4-producing activity and anaerobic digestion efficiency, respectively. The coenzyme F420, unique to hydrogenotrophic methanogens, was correlated with methane yield, suggesting they played a dominant role in the anaerobic reactor. This study demonstrates that the EEM–PARAFAC combined with Q-PCR can be used to characterize methanogenic activity variation during the high-solid anaerobic digestion of rice straw with 15% total solid (TS).