Lignocellulosic biomass is the only renewable carbon resource available in sufficient amount on Earth to go beyond the fossil-based carbon economy. Its transformation requires controlled breakdown of polymers into a set of molecules to make fuels, chemicals and materials. But biomass is a network of various inter-connected polymers which are very difficult to deconstruct optimally. In particular, saccharification potential of lignocellulosic biomass depends on several complex chemical and physical factors. For the first time, an easily measurable fluorescence properties of steam-exploded biomass samples from miscanthus, poplar and wheat straw was shown to be directly correlated to their saccharification potential. Fluorescence can thus be advantageously used as a predictive method of biomass saccharification. The loss in fluorescence occurring after the steam explosion pretreatment and increasing with pretreatment severity does not originate from the loss in lignin content, but rather from a decrease of the lignin β-aryl-ether linkage content. Fluorescence lifetime analysis demonstrates that monolignols making lignin become highly conjugated after steam explosion pretreatment. These results reveal that lignin chemical composition is a more important feature to consider than its content to understand and to predict biomass saccharification.