In this study, we extensively describe experimental models, with correlating experimental conditions, which were used to investigate the enzymatic hydrolysis of bacterial cellulose (BC) to obtain nanocrystals. Cellulase from Trichoderma reesei was used in five enzyme/BC ratios over a period of h. The turbidity data was modeled using both logistic regression and empirical regression to determine the fractal kinetics, resulting in unique kinetic patterns for the mixtures that were richest in BC and in enzymes. The evolution of the yield was inversely related to the turbidity, as confirmed through a semi-empirical approach that was adopted to model the experimental data. The yield values after 74 h of hydrolysis were higher for the substrate-rich mixtures (~20%) than for the enzyme-rich mixtures (~5%), as corroborated by cellobiose and glucose quantification. Transmission electron microscopy and atomic force microscopy analyses revealed a shift from a fibril network to a needlelike morphology (i.e., aggregated nanocrystals or individual nanocrystals ~6 nm width and 200-800 nm in length) as the enzyme/BC ratios went from lower to higher. These results were explained in terms of the heterogeneous substrate model and the erosion model. This work initiated a promising, environmental-friendly method that could serve as an alternative to the commonly used chemical hydrolysis routes.