Methods for the quantification of total and accessible reducing ends on traditional cellulose substrates have been evaluated because of their relevance to enzyme-catalyzed cellulose saccharificaion. For example, quantification of accessible reducing ends is likely to be the most direct measure of substrate concentration for the exo-acting, reducing end-preferring cellobiohydrolases. Two colorimetric assays (dinitrosalicylic acid [DNS] and bicinchoninic acid [BCA] assay) and a radioisotope approach ((NaBH4)-H-3, labeling) were evaluated for this application. Cellulose substrates included microcrystalline celluloses, bacterial celluloses, and filter paper. Estimates of the number of reducing ends per unit mass cellulose were found to be dependent on the assay system (i.e. the DNS and BCA assays gave strikingly different results). DNS-based values were several-fold higher than those obtained using the BCA assay, with fold-differences being substrate specific. Sodium borohydride reduction of celluloses, using cold or radiolabeled reagent under relatively mild conditions, was used to assess the number of surface (solvent-accessible) reducing ends. The results indicate that 30-40% of the reducing ends on traditional cellulose substrates are not solvent accessible; that is, they are buried in the interior of cellulose structures and thus not available to exo-acting enzymes.